Model profile · OpenAI

GPT-5.5

Closed weightsfrontier · registry tag 2026 frontier revision

Verified but aging

Reads as verified but aging across the resolved source data. Capability fit excludes speed and cost.

Visible coverage: 64.4%
Verified coverage: 64.4%
Spread: 83.6%
Last verified: Oct 4, 2026

72%capability fit

textcodevisiondocumentsearch19 aliases47 official source links

Open compare

Data version

Current snapshot.

Published Oct 4, 2026Model list checked4 stale sources105 claim warnings9 providers · 1333 tracked models

The registry snapshot and page stamp are shown so a stale deploy is visible at a glance. View live data status.

Source-linked scores by benchmark

Each row keeps the benchmark source, source type, raw metric, and percentile inside its fair comparison set. Capability summaries exclude archived categories, speed, and cost; those rows remain below.

Verified but agingThis model currently reads as verified but aging across the resolved source data.

Chat / text36 benchmarks74.8%

Intelligence Index

AA · Chat / text · Combined

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #134 · Source label: GPT-5.5 (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 23
Percentile: 76%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `intelligenceIndex`.

76% percentile inside its fair comparison set

23Raw benchmark value

AA-Omniscience accuracy

AA · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #62 · Source label: GPT-5.5 (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 45.6%
Percentile: 86.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `omniscienceAccuracy`.

86.5% percentile inside its fair comparison set

45.6%Raw benchmark value

AA-Omniscience non-hallucination

AA · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #379 · Source label: GPT-5.5 (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 7.3%
Percentile: 16.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `omniscienceNonHallucination`.

16.4% percentile inside its fair comparison set

7.3%Raw benchmark value

IFBench

AA · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #130 · Source label: GPT-5.5 (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 46.1%
Percentile: 62.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `ifbench`.

62.9% percentile inside its fair comparison set

46.1%Raw benchmark value

Input price

AA · Chat / text · Speed / cost

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #324 · Source label: GPT-5.5 (Low)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: $5 /1M input tokens
Percentile: 9.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `price1mInputTokens`.

9.3% percentile inside its fair comparison set

$5 /1M input tokensRaw benchmark value

Output price

AA · Chat / text · Speed / cost

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #332 · Source label: GPT-5.5 (Low)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: $30 /1M output tokens
Percentile: 6.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `price1mOutputTokens`.

6.5% percentile inside its fair comparison set

$30 /1M output tokensRaw benchmark value

Output Speed

AA · Chat / text · Speed / cost

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #137 · Source label: GPT-5.5 (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 81 tokens/s
Percentile: 48.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `medianOutputTokensPerSecond`.

48.1% percentile inside its fair comparison set

81 tokens/sRaw benchmark value

Time to first token

AA · Chat / text · Speed / cost

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #238 · Source label: GPT-5.5 (Xhigh)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 52.31s
Percentile: 10.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `medianTimeToFirstTokenSeconds`.

10.2% percentile inside its fair comparison set

52.31sRaw benchmark value

Time to first answer token

AA · Chat / text · Speed / cost

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #221 · Source label: GPT-5.5 (Xhigh)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 52.31s
Percentile: 16%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `medianTimeToFirstAnswerTokenSeconds`.

16% percentile inside its fair comparison set

52.31sRaw benchmark value

Text Arena

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #25 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,481
Percentile: 93.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: overall. Source rank: #25. Votes: 69202. Organization: openai. License: Proprietary.

93.6% percentile inside its fair comparison set

1,481Raw benchmark valueCI 1,478 - 1,485

Text Arena · Creative Writing

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #34 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,451
Percentile: 91.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: creative_writing. Source rank: #35. Votes: 13120. Organization: openai. License: Proprietary.

91.2% percentile inside its fair comparison set

1,451Raw benchmark valueCI 1,445 - 1,458

Text Arena · English

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #30 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,484
Percentile: 92.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: english. Source rank: #30. Votes: 31442. Organization: openai. License: Proprietary.

92.3% percentile inside its fair comparison set

1,484Raw benchmark valueCI 1,480 - 1,489

Text Arena · Exclude Ties

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #25 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,492
Percentile: 93.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: exclude_ties. Source rank: #25. Votes: 52612. Organization: openai. License: Proprietary.

93.6% percentile inside its fair comparison set

1,492Raw benchmark valueCI 1,487 - 1,497

Text Arena · Hard Prompts

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #31 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,501
Percentile: 92%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: hard_prompts. Source rank: #31. Votes: 45444. Organization: openai. License: Proprietary.

92% percentile inside its fair comparison set

1,501Raw benchmark valueCI 1,497 - 1,506

Text Arena · Hard Prompts English

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #31 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,500
Percentile: 92%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: hard_prompts_english. Source rank: #31. Votes: 21894. Organization: openai. License: Proprietary.

92% percentile inside its fair comparison set

1,500Raw benchmark valueCI 1,495 - 1,506

Text Arena · Instruction Following

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #23 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,478
Percentile: 94.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: instruction_following. Source rank: #23. Votes: 23919. Organization: openai. License: Proprietary.

94.1% percentile inside its fair comparison set

1,478Raw benchmark valueCI 1,472 - 1,483

Text Arena · Longer Query

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #29 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,487
Percentile: 92.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: longer_query. Source rank: #29. Votes: 31949. Organization: openai. License: Proprietary.

92.1% percentile inside its fair comparison set

1,487Raw benchmark valueCI 1,482 - 1,492

Text Arena · Multi Turn

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #26 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,487
Percentile: 93.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: multi_turn. Source rank: #26. Votes: 11824. Organization: openai. License: Proprietary.

93.3% percentile inside its fair comparison set

1,487Raw benchmark valueCI 1,481 - 1,494

Text Arena · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #26 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,471
Percentile: 93.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: overall. Source rank: #27. Votes: 69202. Organization: openai. License: Proprietary.

93.4% percentile inside its fair comparison set

1,471Raw benchmark valueCI 1,468 - 1,475

Text Arena · Creative Writing · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #25 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,455
Percentile: 93.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: creative_writing. Source rank: #26. Votes: 13120. Organization: openai. License: Proprietary.

93.6% percentile inside its fair comparison set

1,455Raw benchmark valueCI 1,448 - 1,462

Text Arena · English · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #36 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,470
Percentile: 90.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: english. Source rank: #37. Votes: 31442. Organization: openai. License: Proprietary.

90.7% percentile inside its fair comparison set

1,470Raw benchmark valueCI 1,465 - 1,475

Text Arena · Exclude Ties · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #27 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,476
Percentile: 93.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: exclude_ties. Source rank: #28. Votes: 52612. Organization: openai. License: Proprietary.

93.1% percentile inside its fair comparison set

1,476Raw benchmark valueCI 1,472 - 1,481

Text Arena · Hard Prompts · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #20 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,489
Percentile: 94.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: hard_prompts. Source rank: #20. Votes: 45444. Organization: openai. License: Proprietary.

94.9% percentile inside its fair comparison set

1,489Raw benchmark valueCI 1,484 - 1,493

Text Arena · Hard Prompts English · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #28 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,484
Percentile: 92.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: hard_prompts_english. Source rank: #28. Votes: 21894. Organization: openai. License: Proprietary.

92.8% percentile inside its fair comparison set

1,484Raw benchmark valueCI 1,478 - 1,489

Text Arena · Instruction Following · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #19 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,478
Percentile: 95.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: instruction_following. Source rank: #19. Votes: 23919. Organization: openai. License: Proprietary.

95.2% percentile inside its fair comparison set

1,478Raw benchmark valueCI 1,473 - 1,484

Text Arena · Longer Query · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #18 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,484
Percentile: 95.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: longer_query. Source rank: #18. Votes: 31949. Organization: openai. License: Proprietary.

95.2% percentile inside its fair comparison set

1,484Raw benchmark valueCI 1,479 - 1,489

Text Arena · Multi Turn · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #27 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,476
Percentile: 93%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: multi_turn. Source rank: #28. Votes: 11824. Organization: openai. License: Proprietary.

93% percentile inside its fair comparison set

1,476Raw benchmark valueCI 1,469 - 1,483

Instruction following

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #27 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 70.7%
Percentile: 60%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Category: IF. Tasks scored: 4.

60% percentile inside its fair comparison set

70.7%Raw benchmark value

Language

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #9 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 87.4%
Percentile: 87.7%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Category: Language. Tasks scored: 3.

87.7% percentile inside its fair comparison set

87.4%Raw benchmark value

Paraphrase

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #21 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 71.7%
Percentile: 69.2%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: paraphrase. Category: IF.

69.2% percentile inside its fair comparison set

71.7%Raw benchmark value

Simplify

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #28 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 66.3%
Percentile: 58.5%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: simplify. Category: IF.

58.5% percentile inside its fair comparison set

66.3%Raw benchmark value

Story generation

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #29 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 72.6%
Percentile: 56.9%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: story_generation. Category: IF.

56.9% percentile inside its fair comparison set

72.6%Raw benchmark value

Summarize

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #26 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 72.3%
Percentile: 61.5%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: summarize. Category: IF.

61.5% percentile inside its fair comparison set

72.3%Raw benchmark value

Connections

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #4 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 100%
Percentile: 100%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: connections. Category: Language.

100% percentile inside its fair comparison set

100%Raw benchmark value

Plot unscrambling

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #12 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 74.1%
Percentile: 83.1%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: plot_unscrambling. Category: Language.

83.1% percentile inside its fair comparison set

74.1%Raw benchmark value

Typos

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #6 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 88%
Percentile: 92.3%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: typos. Category: Language.

92.3% percentile inside its fair comparison set

88%Raw benchmark value

Coding32 benchmarks70.5%

Terminal-Bench Hard

AA · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #8 · Source label: GPT-5.5 (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 49.2%
Percentile: 97.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `terminalbenchHard`.

97.9% percentile inside its fair comparison set

49.2%Raw benchmark value

SciCode

AA · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #47 · Source label: GPT-5.5 (Medium)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 54.5%
Percentile: 76.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `scicode`.

76.8% percentile inside its fair comparison set

54.5%Raw benchmark value

HiL-Bench

SL · Coding · Rubric

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #7 · Source label: GPT-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Scale Labs
Raw value: 39.7%
Percentile: 62.5%
Last updated: recent
Eligibility: headline eligible

Parsed from the public Scale Labs page for scale-hil-bench. Reported model configuration: GPT-5.5. Collapse policy: highest reported score per canonical model.

62.5% percentile inside its fair comparison set

39.7%Raw benchmark value

Text Arena · Coding

AR · Coding · Human

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #40 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,519
Percentile: 89.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: coding. Source rank: #40. Votes: 19110. Organization: openai. License: Proprietary.

89.5% percentile inside its fair comparison set

1,519Raw benchmark valueCI 1,513 - 1,524

Text Arena · Coding · No Style Control

AR · Coding · Human

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #30 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,494
Percentile: 92.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: coding. Source rank: #30. Votes: 19110. Organization: openai. License: Proprietary.

92.2% percentile inside its fair comparison set

1,494Raw benchmark valueCI 1,488 - 1,500

Vibe Code Bench v1.1

VALS-AI · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #33 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 69.8%
Percentile: 65.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: vibe-code; provider: OpenAI.

65.6% percentile inside its fair comparison set

69.8%Raw benchmark valueCI 61% - 78.7%

WebDev Arena

AR · Coding · Human

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #55 · Source label: gpt-5.5-high (codex-harness)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,487
Percentile: 57.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high (codex-harness)`. Category: overall. Source rank: #57. Votes: 18174. Organization: openai. License: Proprietary.

57.5% percentile inside its fair comparison set

1,487Raw benchmark valueCI 1,481 - 1,493

Code Arena · Webdev Html

AR · Coding · Human

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #42 · Source label: gpt-5.5-high (codex-harness)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,535
Percentile: 67.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high (codex-harness)`. Category: webdev-html. Source rank: #43. Votes: 1974. Organization: openai. License: Proprietary.

67.2% percentile inside its fair comparison set

1,535Raw benchmark valueCI 1,521 - 1,549

Code Arena · Webdev React

AR · Coding · Human

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #56 · Source label: gpt-5.5-high (codex-harness)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,478
Percentile: 51.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high (codex-harness)`. Category: webdev-react. Source rank: #58. Votes: 12768. Organization: openai. License: Proprietary.

51.3% percentile inside its fair comparison set

1,478Raw benchmark valueCI 1,471 - 1,484

Code Migration

VALS-AI · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #16 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 45.2%
Percentile: 77.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: code-migration; provider: OpenAI.

77.9% percentile inside its fair comparison set

45.2%Raw benchmark valueCI 37% - 53.3%

ProgramBench

VALS-AI · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #22 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 0.5%
Percentile: 72.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: programbench; provider: OpenAI.

72.4% percentile inside its fair comparison set

0.5%Raw benchmark valueCI 0% - 1.5%

Terminal-Bench 2.1

VALS-AI · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #16 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 76.4%
Percentile: 77.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: terminal-bench-2-1; provider: OpenAI.

77.6% percentile inside its fair comparison set

76.4%Raw benchmark valueCI 75.1% - 77.7%

Code Arena · Image To Webdev

AR · Coding · Human

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #24 · Source label: gpt-5.5-high (codex-harness)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,518
Percentile: 47.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high (codex-harness)`. Category: image_to_webdev. Source rank: #25. Votes: 6877. Organization: openai. License: Proprietary.

47.7% percentile inside its fair comparison set

1,518Raw benchmark valueCI 1,508 - 1,527

SWE-bench Verified

VALS-AI · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #18 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 82.6%
Percentile: 79.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: swebench; provider: OpenAI.

79.3% percentile inside its fair comparison set

82.6%Raw benchmark valueCI 79.3% - 85.9%

LiveCodeBench

VALS-AI · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #34 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 85.3%
Percentile: 73.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: lcb; provider: OpenAI.

73.6% percentile inside its fair comparison set

85.3%Raw benchmark valueCI 83.3% - 87.3%

Agentic coding

LB · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #32 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 54%
Percentile: 52.3%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Category: Agentic Coding. Tasks scored: 3.

52.3% percentile inside its fair comparison set

54%Raw benchmark value

Coding

LB · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #11 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 82.1%
Percentile: 84.6%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Category: Coding. Tasks scored: 2.

84.6% percentile inside its fair comparison set

82.1%Raw benchmark value

JavaScript

LB · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #33 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 63.6%
Percentile: 58.5%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: javascript. Category: Agentic Coding.

58.5% percentile inside its fair comparison set

63.6%Raw benchmark value

TypeScript

LB · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #28 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 43.3%
Percentile: 61.5%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: typescript. Category: Agentic Coding.

61.5% percentile inside its fair comparison set

43.3%Raw benchmark value

Python

LB · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #28 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 55%
Percentile: 63.1%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: python. Category: Agentic Coding.

63.1% percentile inside its fair comparison set

55%Raw benchmark value

Coding generation

LB · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #12 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 81.7%
Percentile: 83.1%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: code_generation. Category: Coding.

83.1% percentile inside its fair comparison set

81.7%Raw benchmark value

Coding completion

LB · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #9 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 82.6%
Percentile: 87.7%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: code_completion. Category: Coding.

87.7% percentile inside its fair comparison set

82.6%Raw benchmark value

Debugging

BB · Coding · Rubric

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #17 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: BridgeBench
Raw value: 77.5%
Percentile: 33.3%
Last updated: archived
Eligibility: headline eligible

Parsed from the BridgeBench page for bridgebench-debugging.

33.3% percentile inside its fair comparison set

77.5%Raw benchmark value

Refactoring

BB · Coding · Rubric

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #11 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: BridgeBench
Raw value: 65.9%
Percentile: 52.4%
Last updated: archived
Eligibility: headline eligible

Parsed from the BridgeBench page for bridgebench-refactoring.

52.4% percentile inside its fair comparison set

65.9%Raw benchmark value

Security

BB · Coding · Rubric

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #1 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: BridgeBench
Raw value: 85.3%
Percentile: 100%
Last updated: archived
Eligibility: headline eligible

Parsed from the BridgeBench page for bridgebench-security.

100% percentile inside its fair comparison set

85.3%Raw benchmark value

Speed throughput

BB · Coding · Speed / cost

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #9 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: BridgeBench
Raw value: 152.3 t/s
Percentile: 68%
Last updated: archived
Eligibility: headline eligible

Parsed from the BridgeBench page for bridgebench-throughput.

68% percentile inside its fair comparison set

152.3 t/sRaw benchmark value

Speed TTFT

BB · Coding · Speed / cost

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #6 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: BridgeBench
Raw value: 930.00ms
Percentile: 80%
Last updated: archived
Eligibility: headline eligible

Parsed from the BridgeBench page for bridgebench-ttft.

80% percentile inside its fair comparison set

930.00msRaw benchmark value

Speed cost

BB · Coding · Speed / cost

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #24 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: BridgeBench
Raw value: $0.9 $
Percentile: 8%
Last updated: archived
Eligibility: headline eligible

Parsed from the BridgeBench page for bridgebench-speed-cost.

8% percentile inside its fair comparison set

$0.9 $Raw benchmark value

Terminal-Bench 2.0

OFF · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #1

Official company resultmanual verifiedmanual verified

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Provider official evals
Raw value: 82.7%
Percentile: 100%
Last updated: archived
Eligibility: headline eligible

Verified against OpenAI's GPT-5.5 launch page dated April 23, 2026. Uses the published Terminal-Bench 2.0 figure. Checked 2026-04-29. Verification: manual_public_page_verification.

100% percentile inside its fair comparison set

82.7%Raw benchmark value

Overall quality

BB · Coding · Combined

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #10 · Source label: GPT-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: BridgeBench
Raw value: 83.6%
Percentile: 59.1%
Last updated: archived
Eligibility: headline eligible

Parsed from the BridgeBench page for bridgebench-overall-quality.

59.1% percentile inside its fair comparison set

83.6%Raw benchmark value

Overall vibe

BB · Coding · Combined

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #8 · Source label: GPT-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: BridgeBench
Raw value: 80%
Percentile: 68.2%
Last updated: archived
Eligibility: headline eligible

Parsed from the BridgeBench page for bridgebench-overall-vibe.

68.2% percentile inside its fair comparison set

80%Raw benchmark value

Overall coverage

BB · Coding · Combined

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #19 · Source label: GPT-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: BridgeBench
Raw value: 85.7%
Percentile: 22.7%
Last updated: archived
Eligibility: headline eligible

Parsed from the BridgeBench page for bridgebench-overall-coverage.

22.7% percentile inside its fair comparison set

85.7%Raw benchmark value

Reasoning / math / science19 benchmarks73.5%

Humanity's Last Exam

AA · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #185 · Source label: GPT-5.5 (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 13.7%
Percentile: 65.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `hle`.

65.2% percentile inside its fair comparison set

13.7%Raw benchmark value

GPQA

AA · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #160 · Source label: GPT-5.5 (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 76.8%
Percentile: 67.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `gpqa`.

67.7% percentile inside its fair comparison set

76.8%Raw benchmark value

CritPt

AA · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #139 · Source label: GPT-5.5 (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 1.4%
Percentile: 69.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `critpt`.

69.7% percentile inside its fair comparison set

1.4%Raw benchmark value

Text Arena · Math

AR · Reasoning / math / science · Human

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #17 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,500
Percentile: 95.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: math. Source rank: #17. Votes: 3499. Organization: openai. License: Proprietary.

95.5% percentile inside its fair comparison set

1,500Raw benchmark valueCI 1,489 - 1,510

Text Arena · Math · No Style Control

AR · Reasoning / math / science · Human

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #25 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,486
Percentile: 93.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: math. Source rank: #26. Votes: 3499. Organization: openai. License: Proprietary.

93.3% percentile inside its fair comparison set

1,486Raw benchmark valueCI 1,476 - 1,497

GPQA Diamond

VALS-AI · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #11 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 93.2%
Percentile: 92.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: gpqa; provider: OpenAI.

92.7% percentile inside its fair comparison set

93.2%Raw benchmark valueCI 90.7% - 95.7%

MMLU Pro

VALS-AI · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #21 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 88.1%
Percentile: 84%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: mmlu_pro; provider: OpenAI.

84% percentile inside its fair comparison set

88.1%Raw benchmark valueCI 87.5% - 88.8%

Mathematics

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #13 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 95.9%
Percentile: 81.5%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Category: Mathematics. Tasks scored: 4.

81.5% percentile inside its fair comparison set

95.9%Raw benchmark value

Reasoning

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #15 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 89.7%
Percentile: 78.5%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Category: Reasoning. Tasks scored: 4.

78.5% percentile inside its fair comparison set

89.7%Raw benchmark value

AMPS Hard

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #48 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 97%
Percentile: 32.3%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: AMPS_Hard. Category: Mathematics.

32.3% percentile inside its fair comparison set

97%Raw benchmark value

Integrals with game

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #6 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 99%
Percentile: 92.3%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: integrals_with_game. Category: Mathematics.

92.3% percentile inside its fair comparison set

99%Raw benchmark value

Math competition

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #26 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 96.1%
Percentile: 66.2%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: math_comp. Category: Mathematics.

66.2% percentile inside its fair comparison set

96.1%Raw benchmark value

Olympiad

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #26 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 91.4%
Percentile: 61.5%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: olympiad. Category: Mathematics.

61.5% percentile inside its fair comparison set

91.4%Raw benchmark value

Theory of mind

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #9 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 84.6%
Percentile: 87.7%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: theory_of_mind. Category: Reasoning.

87.7% percentile inside its fair comparison set

84.6%Raw benchmark value

Zebra puzzle

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #1 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 100%
Percentile: 100%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: zebra_puzzle. Category: Reasoning.

100% percentile inside its fair comparison set

100%Raw benchmark value

Spatial

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #22 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 98%
Percentile: 72.3%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: spatial. Category: Reasoning.

72.3% percentile inside its fair comparison set

98%Raw benchmark value

Logic with navigation

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #24 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 76%
Percentile: 64.6%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: logic_with_navigation. Category: Reasoning.

64.6% percentile inside its fair comparison set

76%Raw benchmark value

Reasoning

BB · Reasoning / math / science · Rubric

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #8 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: BridgeBench
Raw value: 39.7%
Percentile: 66.7%
Last updated: archived
Eligibility: headline eligible

Parsed from the BridgeBench page for bridgebench-reasoning.

66.7% percentile inside its fair comparison set

39.7%Raw benchmark value

Humanity's Last Exam

OFF · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #3

Official company resultmanual verifiedmanual verified

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Provider official evals
Raw value: 41.4%
Percentile: 71.4%
Last updated: archived
Eligibility: headline eligible

Verified against OpenAI's GPT-5.5 launch page dated April 23, 2026. Uses the published no-tools Humanity's Last Exam figure. Checked 2026-04-29. Verification: manual_public_page_verification.

71.4% percentile inside its fair comparison set

41.4%Raw benchmark value

Professional reasoning37 benchmarks77.8%

APEX-Agents-AA

AA · Professional reasoning · Objective

Long-horizon agentic task completion.

Rank #6 · Source label: GPT-5.5 (Xhigh)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 37.7%
Percentile: 84.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `apexAgents`.

84.4% percentile inside its fair comparison set

37.7%Raw benchmark value

Text Arena · Expert

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena expert leaderboard.

Rank #25 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,512
Percentile: 92.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: expert. Source rank: #25. Votes: 7374. Organization: openai. License: Proprietary.

92.7% percentile inside its fair comparison set

1,512Raw benchmark valueCI 1,505 - 1,520

Text Arena · Industry Business And Management And Financial Operations

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_business_and_management_and_financial_operations leaderboard.

Rank #10 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,490
Percentile: 97.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: industry_business_and_management_and_financial_operations. Source rank: #10. Votes: 13704. Organization: openai. License: Proprietary.

97.6% percentile inside its fair comparison set

1,490Raw benchmark valueCI 1,483 - 1,496

Text Arena · Industry Entertainment And Sports And Media

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_entertainment_and_sports_and_media leaderboard.

Rank #34 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,448
Percentile: 91.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: industry_entertainment_and_sports_and_media. Source rank: #35. Votes: 16671. Organization: openai. License: Proprietary.

91.2% percentile inside its fair comparison set

1,448Raw benchmark valueCI 1,442 - 1,454

Text Arena · Industry Legal And Government

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_legal_and_government leaderboard.

Rank #19 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,495
Percentile: 94.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: industry_legal_and_government. Source rank: #19. Votes: 5607. Organization: openai. License: Proprietary.

94.8% percentile inside its fair comparison set

1,495Raw benchmark valueCI 1,486 - 1,504

Text Arena · Industry Life And Physical And Social Science

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_life_and_physical_and_social_science leaderboard.

Rank #25 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,498
Percentile: 93.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: industry_life_and_physical_and_social_science. Source rank: #25. Votes: 11597. Organization: openai. License: Proprietary.

93.6% percentile inside its fair comparison set

1,498Raw benchmark valueCI 1,491 - 1,505

Text Arena · Industry Mathematical

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_mathematical leaderboard.

Rank #14 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,506
Percentile: 96.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: industry_mathematical. Source rank: #14. Votes: 3985. Organization: openai. License: Proprietary.

96.3% percentile inside its fair comparison set

1,506Raw benchmark valueCI 1,496 - 1,516

Text Arena · Industry Medicine And Healthcare

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_medicine_and_healthcare leaderboard.

Rank #40 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,486
Percentile: 88.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: industry_medicine_and_healthcare. Source rank: #41. Votes: 5328. Organization: openai. License: Proprietary.

88.7% percentile inside its fair comparison set

1,486Raw benchmark valueCI 1,477 - 1,495

Text Arena · Industry Software And It Services

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_software_and_it_services leaderboard.

Rank #33 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,510
Percentile: 91.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: industry_software_and_it_services. Source rank: #33. Votes: 27383. Organization: openai. License: Proprietary.

91.5% percentile inside its fair comparison set

1,510Raw benchmark valueCI 1,505 - 1,515

Text Arena · Industry Writing And Literature And Language

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_writing_and_literature_and_language leaderboard.

Rank #24 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,469
Percentile: 93.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: industry_writing_and_literature_and_language. Source rank: #25. Votes: 18108. Organization: openai. License: Proprietary.

93.9% percentile inside its fair comparison set

1,469Raw benchmark valueCI 1,463 - 1,475

Text Arena · Expert · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena expert leaderboard.

Rank #21 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,507
Percentile: 93.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: expert. Source rank: #21. Votes: 7374. Organization: openai. License: Proprietary.

93.9% percentile inside its fair comparison set

1,507Raw benchmark valueCI 1,500 - 1,515

Text Arena · Industry Business And Management And Financial Operations · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_business_and_management_and_financial_operations leaderboard.

Rank #16 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,478
Percentile: 95.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: industry_business_and_management_and_financial_operations. Source rank: #16. Votes: 13704. Organization: openai. License: Proprietary.

95.9% percentile inside its fair comparison set

1,478Raw benchmark valueCI 1,472 - 1,484

Text Arena · Industry Entertainment And Sports And Media · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_entertainment_and_sports_and_media leaderboard.

Rank #28 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,448
Percentile: 92.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: industry_entertainment_and_sports_and_media. Source rank: #29. Votes: 16671. Organization: openai. License: Proprietary.

92.8% percentile inside its fair comparison set

1,448Raw benchmark valueCI 1,442 - 1,454

Text Arena · Industry Legal And Government · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_legal_and_government leaderboard.

Rank #17 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,491
Percentile: 95.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: industry_legal_and_government. Source rank: #17. Votes: 5607. Organization: openai. License: Proprietary.

95.4% percentile inside its fair comparison set

1,491Raw benchmark valueCI 1,482 - 1,500

Text Arena · Industry Life And Physical And Social Science · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_life_and_physical_and_social_science leaderboard.

Rank #26 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,486
Percentile: 93.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: industry_life_and_physical_and_social_science. Source rank: #27. Votes: 11597. Organization: openai. License: Proprietary.

93.3% percentile inside its fair comparison set

1,486Raw benchmark valueCI 1,479 - 1,492

Text Arena · Industry Mathematical · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_mathematical leaderboard.

Rank #21 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,495
Percentile: 94.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: industry_mathematical. Source rank: #21. Votes: 3985. Organization: openai. License: Proprietary.

94.4% percentile inside its fair comparison set

1,495Raw benchmark valueCI 1,485 - 1,505

Text Arena · Industry Medicine And Healthcare · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_medicine_and_healthcare leaderboard.

Rank #44 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,465
Percentile: 87.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: industry_medicine_and_healthcare. Source rank: #45. Votes: 5328. Organization: openai. License: Proprietary.

87.6% percentile inside its fair comparison set

1,465Raw benchmark valueCI 1,456 - 1,474

Text Arena · Industry Software And It Services · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_software_and_it_services leaderboard.

Rank #27 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,492
Percentile: 93.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: industry_software_and_it_services. Source rank: #27. Votes: 27383. Organization: openai. License: Proprietary.

93.1% percentile inside its fair comparison set

1,492Raw benchmark valueCI 1,487 - 1,497

Text Arena · Industry Writing And Literature And Language · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_writing_and_literature_and_language leaderboard.

Rank #23 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,466
Percentile: 94.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: industry_writing_and_literature_and_language. Source rank: #24. Votes: 18108. Organization: openai. License: Proprietary.

94.1% percentile inside its fair comparison set

1,466Raw benchmark valueCI 1,460 - 1,472

Legal Research Bench

VALS-AI · Professional reasoning · Objective

Applied legal research tasks.

Rank #24 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 40.4%
Percentile: 67.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: legal_research; provider: OpenAI.

67.6% percentile inside its fair comparison set

40.4%Raw benchmark valueCI 33.7% - 47.1%

Harvey's Legal Agent Benchmark

VALS-AI · Professional reasoning · Objective

Completing legal work with documents, spreadsheets, presentations, and file-system tools.

Rank #34 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 3.8%
Percentile: 53.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: hlab; provider: OpenAI.

53.6% percentile inside its fair comparison set

3.8%Raw benchmark valueCI 1.4% - 6.1%

LegalBench

VALS-AI · Professional reasoning · Objective

Academic legal reasoning tasks.

Rank #12 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 86.5%
Percentile: 91.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: legal_bench; provider: OpenAI.

91.9% percentile inside its fair comparison set

86.5%Raw benchmark valueCI 85.7% - 87.3%

Finance Agent v2

VALS-AI · Professional reasoning · Objective

Core financial analyst tasks for agentic models.

Rank #30 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 51.8%
Percentile: 58%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: fabv2; provider: OpenAI.

58% percentile inside its fair comparison set

51.8%Raw benchmark valueCI 50.7% - 52.8%

MedCode

VALS-AI · Professional reasoning · Objective

Medical billing support and coding tasks.

Rank #26 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 49.1%
Percentile: 73.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: medcode; provider: OpenAI.

73.7% percentile inside its fair comparison set

49.1%Raw benchmark valueCI 44.8% - 53.4%

MedScribe

VALS-AI · Professional reasoning · Objective

Administrative documentation support for doctors.

Rank #18 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 86.9%
Percentile: 82.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: medscribe; provider: OpenAI.

82.3% percentile inside its fair comparison set

86.9%Raw benchmark valueCI 83.1% - 90.7%

SAGE

VALS-AI · Professional reasoning · Objective

Student Assessment with Generative Evaluation.

Rank #13 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 51.5%
Percentile: 85.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: sage; provider: OpenAI.

85.2% percentile inside its fair comparison set

51.5%Raw benchmark valueCI 43.8% - 59.3%

Public Benefits Bench

VALS-AI · Professional reasoning · Objective

Answering SNAP benefits questions across the public-benefits lifecycle.

Rank #28 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 60.9%
Percentile: 40%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: public-benefits-bench; provider: OpenAI.

40% percentile inside its fair comparison set

60.9%Raw benchmark valueCI 58.4% - 63.4%

SkillsBench

VALS-AI · Professional reasoning · Objective

Applied professional skills tasks.

Rank #4 · Source label: openai/gpt-5.5-codex

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 62.6%
Percentile: 90.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: skillsbench; provider: OpenAI.

90.9% percentile inside its fair comparison set

62.6%Raw benchmark valueCI 53.9% - 71.2%

TaxEval v2

VALS-AI · Professional reasoning · Objective

Answer quality on tax questions and responses.

Rank #25 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 75%
Percentile: 81.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: tax_eval_v2; provider: OpenAI.

81.4% percentile inside its fair comparison set

75%Raw benchmark valueCI 73.3% - 76.7%

Data analysis

LB · Professional reasoning · Objective

Structured data manipulation and table reasoning accuracy.

Rank #4 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 81.6%
Percentile: 95.4%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Category: Data Analysis. Tasks scored: 3.

95.4% percentile inside its fair comparison set

81.6%Raw benchmark value

Overall

LB · Professional reasoning · Objective

Average objective performance across LiveBench's current public category mix.

Rank #11 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 80.2%
Percentile: 84.6%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Category averages included: 7.

84.6% percentile inside its fair comparison set

80.2%Raw benchmark value

Consecutive events

LB · Professional reasoning · Objective

Objective consecutive events score in LiveBench.

Rank #21 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 88.8%
Percentile: 69.2%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: consecutive_events. Category: Data Analysis.

69.2% percentile inside its fair comparison set

88.8%Raw benchmark value

Table join

LB · Professional reasoning · Objective

Objective table join score in LiveBench.

Rank #5 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 55.9%
Percentile: 93.8%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: tablejoin. Category: Data Analysis.

93.8% percentile inside its fair comparison set

55.9%Raw benchmark value

Table reformat

LB · Professional reasoning · Objective

Objective table reformat score in LiveBench.

Rank #6 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 100%
Percentile: 100%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: tablereformat. Category: Data Analysis.

100% percentile inside its fair comparison set

100%Raw benchmark value

Public Benefits Bench v1

VALS-AI · Professional reasoning · Objective

Answering public-benefits questions across the benefits lifecycle.

Rank #8 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 57.2%
Percentile: 41.7%
Last updated: stale
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: public-benefits-bench-v1; provider: OpenAI.

41.7% percentile inside its fair comparison set

57.2%Raw benchmark valueCI 54.7% - 59.8%

Hallucination

BB · Professional reasoning · Rubric

Accuracy and fabrication resistance under prompts that invite unsupported claims.

Rank #13 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: BridgeBench
Raw value: 73.6%
Percentile: 64.7%
Last updated: archived
Eligibility: headline eligible

Parsed from the BridgeBench page for bridgebench-hallucination.

64.7% percentile inside its fair comparison set

73.6%Raw benchmark value

BS pushback

BB · Professional reasoning · Rubric

Resistance to confidently accepting bogus assumptions in expert-style prompts.

Rank #7 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: BridgeBench
Raw value: 88%
Percentile: 72.2%
Last updated: archived
Eligibility: headline eligible

Parsed from the BridgeBench page for bridgebench-pushback.

72.2% percentile inside its fair comparison set

88%Raw benchmark value

Search / tool use4 benchmarks82%

Tau2-Bench Telecom

AA · Search / tool use · Objective

It matters when the model must browse, call tools, and recover useful answers from external systems.

Rank #107 · Source label: GPT-5.5 (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 69.3%
Percentile: 68.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `tau2`.

68.7% percentile inside its fair comparison set

69.3%Raw benchmark value

Search Arena

AR · Search / tool use · Human

It matters when the model must browse, call tools, and recover useful answers from external systems.

Rank #3 · Source label: gpt-5.5-search

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,224
Percentile: 93.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-search`. Category: overall. Source rank: #3. Votes: 89873. Organization: openai. License: Proprietary.

93.9% percentile inside its fair comparison set

1,224Raw benchmark valueCI 1,218 - 1,229

Search Arena · No Style Control

AR · Search / tool use · Human

It matters when the model must browse, call tools, and recover useful answers from external systems.

Rank #3 · Source label: gpt-5.5-search

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,242
Percentile: 93.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-search`. Category: overall. Source rank: #3. Votes: 89873. Organization: openai. License: Proprietary.

93.9% percentile inside its fair comparison set

1,242Raw benchmark valueCI 1,237 - 1,247

BrowseComp

OFF · Search / tool use · Objective

It matters when the model must browse, call tools, and recover useful answers from external systems.

Rank #2

Official company resultmanual verifiedmanual verified

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Provider official evals
Raw value: 84.4%
Percentile: 83.3%
Last updated: archived
Eligibility: headline eligible

Verified against OpenAI's GPT-5.5 launch page dated April 23, 2026. Uses the published BrowseComp figure. Checked 2026-04-29. Verification: manual_public_page_verification.

83.3% percentile inside its fair comparison set

84.4%Raw benchmark value

Long context2 benchmarks77.3%

Long Context Reasoning

AA · Long context · Objective

It checks whether long-context claims survive contact with retrieval, memory, or long-document tasks.

Rank #176 · Source label: GPT-5.5 (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 64%
Percentile: 61.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `lcr`.

61.5% percentile inside its fair comparison set

64%Raw benchmark value

CorpFin v2

VALS-AI · Long context · Objective

It checks whether long-context claims survive contact with retrieval, memory, or long-document tasks.

Rank #9 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 68.4%
Percentile: 93.2%
Last updated: aging
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: corp_fin_v2; provider: OpenAI.

93.2% percentile inside its fair comparison set

68.4%Raw benchmark valueCI 66.6% - 70.2%

Vision understanding19 benchmarks85%

MMMU-Pro

AA · Vision understanding · Objective

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #86 · Source label: GPT-5.5 (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 71.4%
Percentile: 59.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `mmmuPro`.

59.1% percentile inside its fair comparison set

71.4%Raw benchmark value

Vision Arena

AR · Vision understanding · Human

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #16 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,287
Percentile: 89.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: overall. Source rank: #16. Votes: 27223. Organization: openai. License: Proprietary.

89.2% percentile inside its fair comparison set

1,287Raw benchmark valueCI 1,281 - 1,292

Vision Arena · Creative Writing Vision

AR · Vision understanding · Human

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #16 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,285
Percentile: 81.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: creative_writing_vision. Source rank: #16. Votes: 1882. Organization: openai. License: Proprietary.

81.7% percentile inside its fair comparison set

1,285Raw benchmark valueCI 1,270 - 1,299

Vision Arena · Diagram

AR · Vision understanding · Human

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #10 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,324
Percentile: 90.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: diagram. Source rank: #10. Votes: 7401. Organization: openai. License: Proprietary.

90.9% percentile inside its fair comparison set

1,324Raw benchmark valueCI 1,315 - 1,332

Vision Arena · English

AR · Vision understanding · Human

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #15 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,282
Percentile: 89.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: english. Source rank: #15. Votes: 10864. Organization: openai. License: Proprietary.

89.9% percentile inside its fair comparison set

1,282Raw benchmark valueCI 1,274 - 1,291

Vision Arena · Entity Recognition

AR · Vision understanding · Human

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #22 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,213
Percentile: 47.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: entity_recognition. Source rank: #24. Votes: 115. Organization: openai. License: Proprietary.

47.5% percentile inside its fair comparison set

1,213Raw benchmark valueCI 1,159 - 1,267

Vision Arena · Homework

AR · Vision understanding · Human

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #5 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,336
Percentile: 95.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: homework. Source rank: #5. Votes: 3312. Organization: openai. License: Proprietary.

95.8% percentile inside its fair comparison set

1,336Raw benchmark valueCI 1,325 - 1,347

Vision Arena · Humor

AR · Vision understanding · Human

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #12 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,285
Percentile: 84.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: humor. Source rank: #12. Votes: 993. Organization: openai. License: Proprietary.

84.9% percentile inside its fair comparison set

1,285Raw benchmark valueCI 1,266 - 1,305

Vision Arena · Ocr

AR · Vision understanding · Human

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #15 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,301
Percentile: 86%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: ocr. Source rank: #15. Votes: 19233. Organization: openai. License: Proprietary.

86% percentile inside its fair comparison set

1,301Raw benchmark valueCI 1,295 - 1,307

Vision Arena · No Style Control

AR · Vision understanding · Human

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #18 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,297
Percentile: 87.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: overall. Source rank: #19. Votes: 27223. Organization: openai. License: Proprietary.

87.8% percentile inside its fair comparison set

1,297Raw benchmark valueCI 1,291 - 1,303

Vision Arena · Creative Writing Vision · No Style Control

AR · Vision understanding · Human

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #14 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,305
Percentile: 84.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: creative_writing_vision. Source rank: #14. Votes: 1882. Organization: openai. License: Proprietary.

84.1% percentile inside its fair comparison set

1,305Raw benchmark valueCI 1,291 - 1,320

Vision Arena · Diagram · No Style Control

AR · Vision understanding · Human

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #12 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,326
Percentile: 88.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: diagram. Source rank: #12. Votes: 7401. Organization: openai. License: Proprietary.

88.9% percentile inside its fair comparison set

1,326Raw benchmark valueCI 1,317 - 1,334

Vision Arena · English · No Style Control

AR · Vision understanding · Human

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #20 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,293
Percentile: 86.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: english. Source rank: #21. Votes: 10864. Organization: openai. License: Proprietary.

86.2% percentile inside its fair comparison set

1,293Raw benchmark valueCI 1,285 - 1,301

Vision Arena · Entity Recognition · No Style Control

AR · Vision understanding · Human

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #26 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,225
Percentile: 37.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: entity_recognition. Source rank: #30. Votes: 115. Organization: openai. License: Proprietary.

37.5% percentile inside its fair comparison set

1,225Raw benchmark valueCI 1,171 - 1,278

Vision Arena · Homework · No Style Control

AR · Vision understanding · Human

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #14 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,334
Percentile: 86.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: homework. Source rank: #14. Votes: 3312. Organization: openai. License: Proprietary.

86.5% percentile inside its fair comparison set

1,334Raw benchmark valueCI 1,322 - 1,345

Vision Arena · Humor · No Style Control

AR · Vision understanding · Human

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #11 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,296
Percentile: 86.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: humor. Source rank: #12. Votes: 993. Organization: openai. License: Proprietary.

86.3% percentile inside its fair comparison set

1,296Raw benchmark valueCI 1,277 - 1,316

Vision Arena · Ocr · No Style Control

AR · Vision understanding · Human

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #17 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,308
Percentile: 84%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: ocr. Source rank: #18. Votes: 19233. Organization: openai. License: Proprietary.

84% percentile inside its fair comparison set

1,308Raw benchmark valueCI 1,302 - 1,314

MMMU Pro

VALS-AI · Vision understanding · Objective

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #9 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 88.3%
Percentile: 91.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: mmmu; provider: OpenAI.

91.6% percentile inside its fair comparison set

88.3%Raw benchmark valueCI 86.7% - 89.8%

MMMU-Pro

OFF · Vision understanding · Objective

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #1

Official company resultmanual verifiedmanual verified

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Provider official evals
Raw value: 81.2%
Percentile: 100%
Last updated: archived
Eligibility: headline eligible

Verified against OpenAI's GPT-5.5 launch page dated April 23, 2026. Uses the published no-tools MMMU Pro figure. Checked 2026-04-29. Verification: manual_public_page_verification.

100% percentile inside its fair comparison set

81.2%Raw benchmark value

Document understanding4 benchmarks80.7%

Document Arena

AR · Document understanding · Human

It matters when the job is reading PDFs, tables, forms, or mixed-layout documents rather than plain chat.

Rank #10 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,483
Percentile: 77.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: overall. Source rank: #10. Votes: 21279. Organization: openai. License: Proprietary.

77.5% percentile inside its fair comparison set

1,483Raw benchmark valueCI 1,476 - 1,489

Document Arena · No Style Control

AR · Document understanding · Human

It matters when the job is reading PDFs, tables, forms, or mixed-layout documents rather than plain chat.

Rank #8 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,486
Percentile: 82.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: overall. Source rank: #8. Votes: 21279. Organization: openai. License: Proprietary.

82.5% percentile inside its fair comparison set

1,486Raw benchmark valueCI 1,479 - 1,492

MortgageTax

VALS-AI · Document understanding · Objective

It matters when the job is reading PDFs, tables, forms, or mixed-layout documents rather than plain chat.

Rank #10 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 68.8%
Percentile: 89.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: mortgage_tax; provider: OpenAI.

89.5% percentile inside its fair comparison set

68.8%Raw benchmark valueCI 67% - 70.5%

Vals Multimodal Index

VALS-AI · Document understanding · Combined

It matters when the job is reading PDFs, tables, forms, or mixed-layout documents rather than plain chat.

Rank #9 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 68
Percentile: 75%
Last updated: aging
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: vals_multimodal_index; provider: OpenAI.

75% percentile inside its fair comparison set

68Raw benchmark valueCI 65 - 71

Multilingual20 benchmarksn/a

Text Arena · Chinese

AR · Multilingual · Human

Observed user preference in Arena's Text Arena chinese leaderboard.

Rank #25 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,526
Percentile: 93.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: chinese. Source rank: #25. Votes: 4451. Organization: openai. License: Proprietary.

93.2% percentile inside its fair comparison set

1,526Raw benchmark valueCI 1,517 - 1,536

Text Arena · French

AR · Multilingual · Human

Observed user preference in Arena's Text Arena french leaderboard.

Rank #23 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,501
Percentile: 91.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: french. Source rank: #23. Votes: 2508. Organization: openai. License: Proprietary.

91.4% percentile inside its fair comparison set

1,501Raw benchmark valueCI 1,487 - 1,515

Text Arena · German

AR · Multilingual · Human

Observed user preference in Arena's Text Arena german leaderboard.

Rank #16 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,490
Percentile: 94.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: german. Source rank: #16. Votes: 1105. Organization: openai. License: Proprietary.

94.5% percentile inside its fair comparison set

1,490Raw benchmark valueCI 1,471 - 1,509

Text Arena · Japanese

AR · Multilingual · Human

Observed user preference in Arena's Text Arena japanese leaderboard.

Rank #6 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,503
Percentile: 97.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: japanese. Source rank: #6. Votes: 895. Organization: openai. License: Proprietary.

97.9% percentile inside its fair comparison set

1,503Raw benchmark valueCI 1,481 - 1,525

Text Arena · Korean

AR · Multilingual · Human

Observed user preference in Arena's Text Arena korean leaderboard.

Rank #25 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,445
Percentile: 90.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: korean. Source rank: #25. Votes: 1187. Organization: openai. License: Proprietary.

90.2% percentile inside its fair comparison set

1,445Raw benchmark valueCI 1,427 - 1,464

Text Arena · Non English

AR · Multilingual · Human

Observed user preference in Arena's Text Arena non_english leaderboard.

Rank #21 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,474
Percentile: 94.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: non_english. Source rank: #21. Votes: 37759. Organization: openai. License: Proprietary.

94.7% percentile inside its fair comparison set

1,474Raw benchmark valueCI 1,469 - 1,478

Text Arena · Polish

AR · Multilingual · Human

Observed user preference in Arena's Text Arena polish leaderboard.

Rank #29 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,479
Percentile: 85.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: polish. Source rank: #31. Votes: 1172. Organization: openai. License: Proprietary.

85.9% percentile inside its fair comparison set

1,479Raw benchmark valueCI 1,462 - 1,497

Text Arena · Russian

AR · Multilingual · Human

Observed user preference in Arena's Text Arena russian leaderboard.

Rank #31 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,480
Percentile: 91.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: russian. Source rank: #32. Votes: 7339. Organization: openai. License: Proprietary.

91.2% percentile inside its fair comparison set

1,480Raw benchmark valueCI 1,472 - 1,488

Text Arena · Spanish

AR · Multilingual · Human

Observed user preference in Arena's Text Arena spanish leaderboard.

Rank #29 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,469
Percentile: 89.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: spanish. Source rank: #29. Votes: 1937. Organization: openai. License: Proprietary.

89.2% percentile inside its fair comparison set

1,469Raw benchmark valueCI 1,455 - 1,484

Text Arena · Chinese · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena chinese leaderboard.

Rank #12 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,535
Percentile: 96.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: chinese. Source rank: #12. Votes: 4451. Organization: openai. License: Proprietary.

96.9% percentile inside its fair comparison set

1,535Raw benchmark valueCI 1,525 - 1,544

Text Arena · French · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena french leaderboard.

Rank #25 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,486
Percentile: 90.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: french. Source rank: #25. Votes: 2508. Organization: openai. License: Proprietary.

90.7% percentile inside its fair comparison set

1,486Raw benchmark valueCI 1,472 - 1,500

Text Arena · German · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena german leaderboard.

Rank #20 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,481
Percentile: 93%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: german. Source rank: #21. Votes: 1105. Organization: openai. License: Proprietary.

93% percentile inside its fair comparison set

1,481Raw benchmark valueCI 1,462 - 1,500

Text Arena · Japanese · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena japanese leaderboard.

Rank #9 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,497
Percentile: 96.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: japanese. Source rank: #9. Votes: 895. Organization: openai. License: Proprietary.

96.7% percentile inside its fair comparison set

1,497Raw benchmark valueCI 1,475 - 1,519

Text Arena · Korean · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena korean leaderboard.

Rank #12 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,461
Percentile: 95.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: korean. Source rank: #12. Votes: 1187. Organization: openai. License: Proprietary.

95.5% percentile inside its fair comparison set

1,461Raw benchmark valueCI 1,443 - 1,479

Text Arena · Non English · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena non_english leaderboard.

Rank #23 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,467
Percentile: 94.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: non_english. Source rank: #24. Votes: 37759. Organization: openai. License: Proprietary.

94.1% percentile inside its fair comparison set

1,467Raw benchmark valueCI 1,462 - 1,471

Text Arena · Polish · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena polish leaderboard.

Rank #29 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,465
Percentile: 85.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: polish. Source rank: #31. Votes: 1172. Organization: openai. License: Proprietary.

85.9% percentile inside its fair comparison set

1,465Raw benchmark valueCI 1,448 - 1,483

Text Arena · Russian · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena russian leaderboard.

Rank #27 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,473
Percentile: 92.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: russian. Source rank: #28. Votes: 7403. Organization: openai. License: Proprietary.

92.4% percentile inside its fair comparison set

1,473Raw benchmark valueCI 1,465 - 1,481

Text Arena · Spanish · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena spanish leaderboard.

Rank #30 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,468
Percentile: 88.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: spanish. Source rank: #31. Votes: 1937. Organization: openai. License: Proprietary.

88.8% percentile inside its fair comparison set

1,468Raw benchmark valueCI 1,453 - 1,482

Vision Arena · Chinese

AR · Multilingual · Human

Observed user preference in Arena's Vision Arena chinese leaderboard.

Rank #21 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,328
Percentile: 80.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: chinese. Source rank: #22. Votes: 1711. Organization: openai. License: Proprietary.

80.8% percentile inside its fair comparison set

1,328Raw benchmark valueCI 1,311 - 1,345

Vision Arena · Chinese · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Vision Arena chinese leaderboard.

Rank #15 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,359
Percentile: 86.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: chinese. Source rank: #16. Votes: 1502. Organization: openai. License: Proprietary.

86.5% percentile inside its fair comparison set

1,359Raw benchmark valueCI 1,341 - 1,377

Source links and registry checks

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Artificial Analysis

Oct 4, 2026

source →

official

LiveBench

Oct 4, 2026

source →

official

Scale Labs

Oct 4, 2026

source →

Model profile · OpenAI

GPT-5.5

Closed weightsfrontier · registry tag 2026 frontier revision

Verified but aging

Reads as verified but aging across the resolved source data. Capability fit excludes speed and cost.

Visible coverage: 64.4%
Verified coverage: 64.4%
Spread: 83.6%
Last verified: Oct 4, 2026

72%capability fit

textcodevisiondocumentsearch19 aliases47 official source links

Open compare

Data version

Current snapshot.

Published Oct 4, 2026Model list checked4 stale sources105 claim warnings9 providers · 1333 tracked models

The registry snapshot and page stamp are shown so a stale deploy is visible at a glance. View live data status.

Source-linked scores by benchmark

Verified but agingThis model currently reads as verified but aging across the resolved source data.

Chat / text36 benchmarks74.8%

Intelligence Index

AA · Chat / text · Combined

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #134 · Source label: GPT-5.5 (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 23
Percentile: 76%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `intelligenceIndex`.

76% percentile inside its fair comparison set

23Raw benchmark value

AA-Omniscience accuracy

AA · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #62 · Source label: GPT-5.5 (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 45.6%
Percentile: 86.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `omniscienceAccuracy`.

86.5% percentile inside its fair comparison set

45.6%Raw benchmark value

AA-Omniscience non-hallucination

AA · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #379 · Source label: GPT-5.5 (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 7.3%
Percentile: 16.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `omniscienceNonHallucination`.

16.4% percentile inside its fair comparison set

7.3%Raw benchmark value

IFBench

AA · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #130 · Source label: GPT-5.5 (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 46.1%
Percentile: 62.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `ifbench`.

62.9% percentile inside its fair comparison set

46.1%Raw benchmark value

Input price

AA · Chat / text · Speed / cost

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #324 · Source label: GPT-5.5 (Low)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: $5 /1M input tokens
Percentile: 9.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `price1mInputTokens`.

9.3% percentile inside its fair comparison set

$5 /1M input tokensRaw benchmark value

Output price

AA · Chat / text · Speed / cost

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #332 · Source label: GPT-5.5 (Low)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: $30 /1M output tokens
Percentile: 6.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `price1mOutputTokens`.

6.5% percentile inside its fair comparison set

$30 /1M output tokensRaw benchmark value

Output Speed

AA · Chat / text · Speed / cost

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #137 · Source label: GPT-5.5 (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 81 tokens/s
Percentile: 48.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `medianOutputTokensPerSecond`.

48.1% percentile inside its fair comparison set

81 tokens/sRaw benchmark value

Time to first token

AA · Chat / text · Speed / cost

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #238 · Source label: GPT-5.5 (Xhigh)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 52.31s
Percentile: 10.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `medianTimeToFirstTokenSeconds`.

10.2% percentile inside its fair comparison set

52.31sRaw benchmark value

Time to first answer token

AA · Chat / text · Speed / cost

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #221 · Source label: GPT-5.5 (Xhigh)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 52.31s
Percentile: 16%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `medianTimeToFirstAnswerTokenSeconds`.

16% percentile inside its fair comparison set

52.31sRaw benchmark value

Text Arena

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #25 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,481
Percentile: 93.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: overall. Source rank: #25. Votes: 69202. Organization: openai. License: Proprietary.

93.6% percentile inside its fair comparison set

1,481Raw benchmark valueCI 1,478 - 1,485

Text Arena · Creative Writing

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #34 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,451
Percentile: 91.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: creative_writing. Source rank: #35. Votes: 13120. Organization: openai. License: Proprietary.

91.2% percentile inside its fair comparison set

1,451Raw benchmark valueCI 1,445 - 1,458

Text Arena · English

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #30 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,484
Percentile: 92.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: english. Source rank: #30. Votes: 31442. Organization: openai. License: Proprietary.

92.3% percentile inside its fair comparison set

1,484Raw benchmark valueCI 1,480 - 1,489

Text Arena · Exclude Ties

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #25 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,492
Percentile: 93.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: exclude_ties. Source rank: #25. Votes: 52612. Organization: openai. License: Proprietary.

93.6% percentile inside its fair comparison set

1,492Raw benchmark valueCI 1,487 - 1,497

Text Arena · Hard Prompts

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #31 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,501
Percentile: 92%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: hard_prompts. Source rank: #31. Votes: 45444. Organization: openai. License: Proprietary.

92% percentile inside its fair comparison set

1,501Raw benchmark valueCI 1,497 - 1,506

Text Arena · Hard Prompts English

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #31 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,500
Percentile: 92%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: hard_prompts_english. Source rank: #31. Votes: 21894. Organization: openai. License: Proprietary.

92% percentile inside its fair comparison set

1,500Raw benchmark valueCI 1,495 - 1,506

Text Arena · Instruction Following

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #23 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,478
Percentile: 94.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: instruction_following. Source rank: #23. Votes: 23919. Organization: openai. License: Proprietary.

94.1% percentile inside its fair comparison set

1,478Raw benchmark valueCI 1,472 - 1,483

Text Arena · Longer Query

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #29 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,487
Percentile: 92.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: longer_query. Source rank: #29. Votes: 31949. Organization: openai. License: Proprietary.

92.1% percentile inside its fair comparison set

1,487Raw benchmark valueCI 1,482 - 1,492

Text Arena · Multi Turn

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #26 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,487
Percentile: 93.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: multi_turn. Source rank: #26. Votes: 11824. Organization: openai. License: Proprietary.

93.3% percentile inside its fair comparison set

1,487Raw benchmark valueCI 1,481 - 1,494

Text Arena · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #26 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,471
Percentile: 93.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: overall. Source rank: #27. Votes: 69202. Organization: openai. License: Proprietary.

93.4% percentile inside its fair comparison set

1,471Raw benchmark valueCI 1,468 - 1,475

Text Arena · Creative Writing · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #25 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,455
Percentile: 93.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: creative_writing. Source rank: #26. Votes: 13120. Organization: openai. License: Proprietary.

93.6% percentile inside its fair comparison set

1,455Raw benchmark valueCI 1,448 - 1,462

Text Arena · English · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #36 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,470
Percentile: 90.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: english. Source rank: #37. Votes: 31442. Organization: openai. License: Proprietary.

90.7% percentile inside its fair comparison set

1,470Raw benchmark valueCI 1,465 - 1,475

Text Arena · Exclude Ties · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #27 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,476
Percentile: 93.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: exclude_ties. Source rank: #28. Votes: 52612. Organization: openai. License: Proprietary.

93.1% percentile inside its fair comparison set

1,476Raw benchmark valueCI 1,472 - 1,481

Text Arena · Hard Prompts · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #20 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,489
Percentile: 94.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: hard_prompts. Source rank: #20. Votes: 45444. Organization: openai. License: Proprietary.

94.9% percentile inside its fair comparison set

1,489Raw benchmark valueCI 1,484 - 1,493

Text Arena · Hard Prompts English · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #28 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,484
Percentile: 92.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: hard_prompts_english. Source rank: #28. Votes: 21894. Organization: openai. License: Proprietary.

92.8% percentile inside its fair comparison set

1,484Raw benchmark valueCI 1,478 - 1,489

Text Arena · Instruction Following · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #19 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,478
Percentile: 95.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: instruction_following. Source rank: #19. Votes: 23919. Organization: openai. License: Proprietary.

95.2% percentile inside its fair comparison set

1,478Raw benchmark valueCI 1,473 - 1,484

Text Arena · Longer Query · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #18 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,484
Percentile: 95.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: longer_query. Source rank: #18. Votes: 31949. Organization: openai. License: Proprietary.

95.2% percentile inside its fair comparison set

1,484Raw benchmark valueCI 1,479 - 1,489

Text Arena · Multi Turn · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #27 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,476
Percentile: 93%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: multi_turn. Source rank: #28. Votes: 11824. Organization: openai. License: Proprietary.

93% percentile inside its fair comparison set

1,476Raw benchmark valueCI 1,469 - 1,483

Instruction following

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #27 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 70.7%
Percentile: 60%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Category: IF. Tasks scored: 4.

60% percentile inside its fair comparison set

70.7%Raw benchmark value

Language

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #9 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 87.4%
Percentile: 87.7%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Category: Language. Tasks scored: 3.

87.7% percentile inside its fair comparison set

87.4%Raw benchmark value

Paraphrase

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #21 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 71.7%
Percentile: 69.2%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: paraphrase. Category: IF.

69.2% percentile inside its fair comparison set

71.7%Raw benchmark value

Simplify

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #28 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 66.3%
Percentile: 58.5%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: simplify. Category: IF.

58.5% percentile inside its fair comparison set

66.3%Raw benchmark value

Story generation

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #29 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 72.6%
Percentile: 56.9%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: story_generation. Category: IF.

56.9% percentile inside its fair comparison set

72.6%Raw benchmark value

Summarize

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #26 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 72.3%
Percentile: 61.5%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: summarize. Category: IF.

61.5% percentile inside its fair comparison set

72.3%Raw benchmark value

Connections

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #4 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 100%
Percentile: 100%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: connections. Category: Language.

100% percentile inside its fair comparison set

100%Raw benchmark value

Plot unscrambling

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #12 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 74.1%
Percentile: 83.1%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: plot_unscrambling. Category: Language.

83.1% percentile inside its fair comparison set

74.1%Raw benchmark value

Typos

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #6 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 88%
Percentile: 92.3%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: typos. Category: Language.

92.3% percentile inside its fair comparison set

88%Raw benchmark value

Coding32 benchmarks70.5%

Terminal-Bench Hard

AA · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #8 · Source label: GPT-5.5 (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 49.2%
Percentile: 97.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `terminalbenchHard`.

97.9% percentile inside its fair comparison set

49.2%Raw benchmark value

SciCode

AA · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #47 · Source label: GPT-5.5 (Medium)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 54.5%
Percentile: 76.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `scicode`.

76.8% percentile inside its fair comparison set

54.5%Raw benchmark value

HiL-Bench

SL · Coding · Rubric

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #7 · Source label: GPT-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Scale Labs
Raw value: 39.7%
Percentile: 62.5%
Last updated: recent
Eligibility: headline eligible

Parsed from the public Scale Labs page for scale-hil-bench. Reported model configuration: GPT-5.5. Collapse policy: highest reported score per canonical model.

62.5% percentile inside its fair comparison set

39.7%Raw benchmark value

Text Arena · Coding

AR · Coding · Human

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #40 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,519
Percentile: 89.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: coding. Source rank: #40. Votes: 19110. Organization: openai. License: Proprietary.

89.5% percentile inside its fair comparison set

1,519Raw benchmark valueCI 1,513 - 1,524

Text Arena · Coding · No Style Control

AR · Coding · Human

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #30 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,494
Percentile: 92.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: coding. Source rank: #30. Votes: 19110. Organization: openai. License: Proprietary.

92.2% percentile inside its fair comparison set

1,494Raw benchmark valueCI 1,488 - 1,500

Vibe Code Bench v1.1

VALS-AI · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #33 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 69.8%
Percentile: 65.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: vibe-code; provider: OpenAI.

65.6% percentile inside its fair comparison set

69.8%Raw benchmark valueCI 61% - 78.7%

WebDev Arena

AR · Coding · Human

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #55 · Source label: gpt-5.5-high (codex-harness)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,487
Percentile: 57.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high (codex-harness)`. Category: overall. Source rank: #57. Votes: 18174. Organization: openai. License: Proprietary.

57.5% percentile inside its fair comparison set

1,487Raw benchmark valueCI 1,481 - 1,493

Code Arena · Webdev Html

AR · Coding · Human

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #42 · Source label: gpt-5.5-high (codex-harness)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,535
Percentile: 67.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high (codex-harness)`. Category: webdev-html. Source rank: #43. Votes: 1974. Organization: openai. License: Proprietary.

67.2% percentile inside its fair comparison set

1,535Raw benchmark valueCI 1,521 - 1,549

Code Arena · Webdev React

AR · Coding · Human

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #56 · Source label: gpt-5.5-high (codex-harness)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,478
Percentile: 51.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high (codex-harness)`. Category: webdev-react. Source rank: #58. Votes: 12768. Organization: openai. License: Proprietary.

51.3% percentile inside its fair comparison set

1,478Raw benchmark valueCI 1,471 - 1,484

Code Migration

VALS-AI · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #16 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 45.2%
Percentile: 77.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: code-migration; provider: OpenAI.

77.9% percentile inside its fair comparison set

45.2%Raw benchmark valueCI 37% - 53.3%

ProgramBench

VALS-AI · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #22 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 0.5%
Percentile: 72.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: programbench; provider: OpenAI.

72.4% percentile inside its fair comparison set

0.5%Raw benchmark valueCI 0% - 1.5%

Terminal-Bench 2.1

VALS-AI · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #16 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 76.4%
Percentile: 77.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: terminal-bench-2-1; provider: OpenAI.

77.6% percentile inside its fair comparison set

76.4%Raw benchmark valueCI 75.1% - 77.7%

Code Arena · Image To Webdev

AR · Coding · Human

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #24 · Source label: gpt-5.5-high (codex-harness)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,518
Percentile: 47.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high (codex-harness)`. Category: image_to_webdev. Source rank: #25. Votes: 6877. Organization: openai. License: Proprietary.

47.7% percentile inside its fair comparison set

1,518Raw benchmark valueCI 1,508 - 1,527

SWE-bench Verified

VALS-AI · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #18 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 82.6%
Percentile: 79.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: swebench; provider: OpenAI.

79.3% percentile inside its fair comparison set

82.6%Raw benchmark valueCI 79.3% - 85.9%

LiveCodeBench

VALS-AI · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #34 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 85.3%
Percentile: 73.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: lcb; provider: OpenAI.

73.6% percentile inside its fair comparison set

85.3%Raw benchmark valueCI 83.3% - 87.3%

Agentic coding

LB · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #32 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 54%
Percentile: 52.3%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Category: Agentic Coding. Tasks scored: 3.

52.3% percentile inside its fair comparison set

54%Raw benchmark value

Coding

LB · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #11 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 82.1%
Percentile: 84.6%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Category: Coding. Tasks scored: 2.

84.6% percentile inside its fair comparison set

82.1%Raw benchmark value

JavaScript

LB · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #33 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 63.6%
Percentile: 58.5%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: javascript. Category: Agentic Coding.

58.5% percentile inside its fair comparison set

63.6%Raw benchmark value

TypeScript

LB · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #28 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 43.3%
Percentile: 61.5%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: typescript. Category: Agentic Coding.

61.5% percentile inside its fair comparison set

43.3%Raw benchmark value

Python

LB · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #28 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 55%
Percentile: 63.1%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: python. Category: Agentic Coding.

63.1% percentile inside its fair comparison set

55%Raw benchmark value

Coding generation

LB · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #12 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 81.7%
Percentile: 83.1%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: code_generation. Category: Coding.

83.1% percentile inside its fair comparison set

81.7%Raw benchmark value

Coding completion

LB · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #9 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 82.6%
Percentile: 87.7%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: code_completion. Category: Coding.

87.7% percentile inside its fair comparison set

82.6%Raw benchmark value

Debugging

BB · Coding · Rubric

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #17 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: BridgeBench
Raw value: 77.5%
Percentile: 33.3%
Last updated: archived
Eligibility: headline eligible

Parsed from the BridgeBench page for bridgebench-debugging.

33.3% percentile inside its fair comparison set

77.5%Raw benchmark value

Refactoring

BB · Coding · Rubric

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #11 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: BridgeBench
Raw value: 65.9%
Percentile: 52.4%
Last updated: archived
Eligibility: headline eligible

Parsed from the BridgeBench page for bridgebench-refactoring.

52.4% percentile inside its fair comparison set

65.9%Raw benchmark value

Security

BB · Coding · Rubric

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #1 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: BridgeBench
Raw value: 85.3%
Percentile: 100%
Last updated: archived
Eligibility: headline eligible

Parsed from the BridgeBench page for bridgebench-security.

100% percentile inside its fair comparison set

85.3%Raw benchmark value

Speed throughput

BB · Coding · Speed / cost

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #9 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: BridgeBench
Raw value: 152.3 t/s
Percentile: 68%
Last updated: archived
Eligibility: headline eligible

Parsed from the BridgeBench page for bridgebench-throughput.

68% percentile inside its fair comparison set

152.3 t/sRaw benchmark value

Speed TTFT

BB · Coding · Speed / cost

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #6 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: BridgeBench
Raw value: 930.00ms
Percentile: 80%
Last updated: archived
Eligibility: headline eligible

Parsed from the BridgeBench page for bridgebench-ttft.

80% percentile inside its fair comparison set

930.00msRaw benchmark value

Speed cost

BB · Coding · Speed / cost

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #24 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: BridgeBench
Raw value: $0.9 $
Percentile: 8%
Last updated: archived
Eligibility: headline eligible

Parsed from the BridgeBench page for bridgebench-speed-cost.

8% percentile inside its fair comparison set

$0.9 $Raw benchmark value

Terminal-Bench 2.0

OFF · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #1

Official company resultmanual verifiedmanual verified

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Provider official evals
Raw value: 82.7%
Percentile: 100%
Last updated: archived
Eligibility: headline eligible

Verified against OpenAI's GPT-5.5 launch page dated April 23, 2026. Uses the published Terminal-Bench 2.0 figure. Checked 2026-04-29. Verification: manual_public_page_verification.

100% percentile inside its fair comparison set

82.7%Raw benchmark value

Overall quality

BB · Coding · Combined

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #10 · Source label: GPT-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: BridgeBench
Raw value: 83.6%
Percentile: 59.1%
Last updated: archived
Eligibility: headline eligible

Parsed from the BridgeBench page for bridgebench-overall-quality.

59.1% percentile inside its fair comparison set

83.6%Raw benchmark value

Overall vibe

BB · Coding · Combined

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #8 · Source label: GPT-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: BridgeBench
Raw value: 80%
Percentile: 68.2%
Last updated: archived
Eligibility: headline eligible

Parsed from the BridgeBench page for bridgebench-overall-vibe.

68.2% percentile inside its fair comparison set

80%Raw benchmark value

Overall coverage

BB · Coding · Combined

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #19 · Source label: GPT-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: BridgeBench
Raw value: 85.7%
Percentile: 22.7%
Last updated: archived
Eligibility: headline eligible

Parsed from the BridgeBench page for bridgebench-overall-coverage.

22.7% percentile inside its fair comparison set

85.7%Raw benchmark value

Reasoning / math / science19 benchmarks73.5%

Humanity's Last Exam

AA · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #185 · Source label: GPT-5.5 (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 13.7%
Percentile: 65.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `hle`.

65.2% percentile inside its fair comparison set

13.7%Raw benchmark value

GPQA

AA · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #160 · Source label: GPT-5.5 (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 76.8%
Percentile: 67.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `gpqa`.

67.7% percentile inside its fair comparison set

76.8%Raw benchmark value

CritPt

AA · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #139 · Source label: GPT-5.5 (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 1.4%
Percentile: 69.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `critpt`.

69.7% percentile inside its fair comparison set

1.4%Raw benchmark value

Text Arena · Math

AR · Reasoning / math / science · Human

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #17 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,500
Percentile: 95.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: math. Source rank: #17. Votes: 3499. Organization: openai. License: Proprietary.

95.5% percentile inside its fair comparison set

1,500Raw benchmark valueCI 1,489 - 1,510

Text Arena · Math · No Style Control

AR · Reasoning / math / science · Human

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #25 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,486
Percentile: 93.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: math. Source rank: #26. Votes: 3499. Organization: openai. License: Proprietary.

93.3% percentile inside its fair comparison set

1,486Raw benchmark valueCI 1,476 - 1,497

GPQA Diamond

VALS-AI · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #11 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 93.2%
Percentile: 92.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: gpqa; provider: OpenAI.

92.7% percentile inside its fair comparison set

93.2%Raw benchmark valueCI 90.7% - 95.7%

MMLU Pro

VALS-AI · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #21 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 88.1%
Percentile: 84%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: mmlu_pro; provider: OpenAI.

84% percentile inside its fair comparison set

88.1%Raw benchmark valueCI 87.5% - 88.8%

Mathematics

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #13 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 95.9%
Percentile: 81.5%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Category: Mathematics. Tasks scored: 4.

81.5% percentile inside its fair comparison set

95.9%Raw benchmark value

Reasoning

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #15 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 89.7%
Percentile: 78.5%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Category: Reasoning. Tasks scored: 4.

78.5% percentile inside its fair comparison set

89.7%Raw benchmark value

AMPS Hard

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #48 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 97%
Percentile: 32.3%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: AMPS_Hard. Category: Mathematics.

32.3% percentile inside its fair comparison set

97%Raw benchmark value

Integrals with game

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #6 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 99%
Percentile: 92.3%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: integrals_with_game. Category: Mathematics.

92.3% percentile inside its fair comparison set

99%Raw benchmark value

Math competition

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #26 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 96.1%
Percentile: 66.2%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: math_comp. Category: Mathematics.

66.2% percentile inside its fair comparison set

96.1%Raw benchmark value

Olympiad

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #26 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 91.4%
Percentile: 61.5%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: olympiad. Category: Mathematics.

61.5% percentile inside its fair comparison set

91.4%Raw benchmark value

Theory of mind

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #9 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 84.6%
Percentile: 87.7%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: theory_of_mind. Category: Reasoning.

87.7% percentile inside its fair comparison set

84.6%Raw benchmark value

Zebra puzzle

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #1 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 100%
Percentile: 100%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: zebra_puzzle. Category: Reasoning.

100% percentile inside its fair comparison set

100%Raw benchmark value

Spatial

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #22 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 98%
Percentile: 72.3%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: spatial. Category: Reasoning.

72.3% percentile inside its fair comparison set

98%Raw benchmark value

Logic with navigation

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #24 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 76%
Percentile: 64.6%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: logic_with_navigation. Category: Reasoning.

64.6% percentile inside its fair comparison set

76%Raw benchmark value

Reasoning

BB · Reasoning / math / science · Rubric

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #8 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: BridgeBench
Raw value: 39.7%
Percentile: 66.7%
Last updated: archived
Eligibility: headline eligible

Parsed from the BridgeBench page for bridgebench-reasoning.

66.7% percentile inside its fair comparison set

39.7%Raw benchmark value

Humanity's Last Exam

OFF · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #3

Official company resultmanual verifiedmanual verified

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Provider official evals
Raw value: 41.4%
Percentile: 71.4%
Last updated: archived
Eligibility: headline eligible

Verified against OpenAI's GPT-5.5 launch page dated April 23, 2026. Uses the published no-tools Humanity's Last Exam figure. Checked 2026-04-29. Verification: manual_public_page_verification.

71.4% percentile inside its fair comparison set

41.4%Raw benchmark value

Professional reasoning37 benchmarks77.8%

APEX-Agents-AA

AA · Professional reasoning · Objective

Long-horizon agentic task completion.

Rank #6 · Source label: GPT-5.5 (Xhigh)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 37.7%
Percentile: 84.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `apexAgents`.

84.4% percentile inside its fair comparison set

37.7%Raw benchmark value

Text Arena · Expert

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena expert leaderboard.

Rank #25 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,512
Percentile: 92.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: expert. Source rank: #25. Votes: 7374. Organization: openai. License: Proprietary.

92.7% percentile inside its fair comparison set

1,512Raw benchmark valueCI 1,505 - 1,520

Text Arena · Industry Business And Management And Financial Operations

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_business_and_management_and_financial_operations leaderboard.

Rank #10 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,490
Percentile: 97.6%
Last updated: recent
Eligibility: headline eligible

97.6% percentile inside its fair comparison set

1,490Raw benchmark valueCI 1,483 - 1,496

Text Arena · Industry Entertainment And Sports And Media

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_entertainment_and_sports_and_media leaderboard.

Rank #34 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,448
Percentile: 91.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: industry_entertainment_and_sports_and_media. Source rank: #35. Votes: 16671. Organization: openai. License: Proprietary.

91.2% percentile inside its fair comparison set

1,448Raw benchmark valueCI 1,442 - 1,454

Text Arena · Industry Legal And Government

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_legal_and_government leaderboard.

Rank #19 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,495
Percentile: 94.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: industry_legal_and_government. Source rank: #19. Votes: 5607. Organization: openai. License: Proprietary.

94.8% percentile inside its fair comparison set

1,495Raw benchmark valueCI 1,486 - 1,504

Text Arena · Industry Life And Physical And Social Science

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_life_and_physical_and_social_science leaderboard.

Rank #25 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,498
Percentile: 93.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: industry_life_and_physical_and_social_science. Source rank: #25. Votes: 11597. Organization: openai. License: Proprietary.

93.6% percentile inside its fair comparison set

1,498Raw benchmark valueCI 1,491 - 1,505

Text Arena · Industry Mathematical

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_mathematical leaderboard.

Rank #14 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,506
Percentile: 96.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: industry_mathematical. Source rank: #14. Votes: 3985. Organization: openai. License: Proprietary.

96.3% percentile inside its fair comparison set

1,506Raw benchmark valueCI 1,496 - 1,516

Text Arena · Industry Medicine And Healthcare

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_medicine_and_healthcare leaderboard.

Rank #40 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,486
Percentile: 88.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: industry_medicine_and_healthcare. Source rank: #41. Votes: 5328. Organization: openai. License: Proprietary.

88.7% percentile inside its fair comparison set

1,486Raw benchmark valueCI 1,477 - 1,495

Text Arena · Industry Software And It Services

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_software_and_it_services leaderboard.

Rank #33 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,510
Percentile: 91.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: industry_software_and_it_services. Source rank: #33. Votes: 27383. Organization: openai. License: Proprietary.

91.5% percentile inside its fair comparison set

1,510Raw benchmark valueCI 1,505 - 1,515

Text Arena · Industry Writing And Literature And Language

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_writing_and_literature_and_language leaderboard.

Rank #24 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,469
Percentile: 93.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: industry_writing_and_literature_and_language. Source rank: #25. Votes: 18108. Organization: openai. License: Proprietary.

93.9% percentile inside its fair comparison set

1,469Raw benchmark valueCI 1,463 - 1,475

Text Arena · Expert · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena expert leaderboard.

Rank #21 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,507
Percentile: 93.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: expert. Source rank: #21. Votes: 7374. Organization: openai. License: Proprietary.

93.9% percentile inside its fair comparison set

1,507Raw benchmark valueCI 1,500 - 1,515

Text Arena · Industry Business And Management And Financial Operations · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_business_and_management_and_financial_operations leaderboard.

Rank #16 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,478
Percentile: 95.9%
Last updated: recent
Eligibility: headline eligible

95.9% percentile inside its fair comparison set

1,478Raw benchmark valueCI 1,472 - 1,484

Text Arena · Industry Entertainment And Sports And Media · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_entertainment_and_sports_and_media leaderboard.

Rank #28 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,448
Percentile: 92.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: industry_entertainment_and_sports_and_media. Source rank: #29. Votes: 16671. Organization: openai. License: Proprietary.

92.8% percentile inside its fair comparison set

1,448Raw benchmark valueCI 1,442 - 1,454

Text Arena · Industry Legal And Government · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_legal_and_government leaderboard.

Rank #17 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,491
Percentile: 95.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: industry_legal_and_government. Source rank: #17. Votes: 5607. Organization: openai. License: Proprietary.

95.4% percentile inside its fair comparison set

1,491Raw benchmark valueCI 1,482 - 1,500

Text Arena · Industry Life And Physical And Social Science · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_life_and_physical_and_social_science leaderboard.

Rank #26 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,486
Percentile: 93.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: industry_life_and_physical_and_social_science. Source rank: #27. Votes: 11597. Organization: openai. License: Proprietary.

93.3% percentile inside its fair comparison set

1,486Raw benchmark valueCI 1,479 - 1,492

Text Arena · Industry Mathematical · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_mathematical leaderboard.

Rank #21 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,495
Percentile: 94.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: industry_mathematical. Source rank: #21. Votes: 3985. Organization: openai. License: Proprietary.

94.4% percentile inside its fair comparison set

1,495Raw benchmark valueCI 1,485 - 1,505

Text Arena · Industry Medicine And Healthcare · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_medicine_and_healthcare leaderboard.

Rank #44 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,465
Percentile: 87.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: industry_medicine_and_healthcare. Source rank: #45. Votes: 5328. Organization: openai. License: Proprietary.

87.6% percentile inside its fair comparison set

1,465Raw benchmark valueCI 1,456 - 1,474

Text Arena · Industry Software And It Services · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_software_and_it_services leaderboard.

Rank #27 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,492
Percentile: 93.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: industry_software_and_it_services. Source rank: #27. Votes: 27383. Organization: openai. License: Proprietary.

93.1% percentile inside its fair comparison set

1,492Raw benchmark valueCI 1,487 - 1,497

Text Arena · Industry Writing And Literature And Language · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_writing_and_literature_and_language leaderboard.

Rank #23 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,466
Percentile: 94.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: industry_writing_and_literature_and_language. Source rank: #24. Votes: 18108. Organization: openai. License: Proprietary.

94.1% percentile inside its fair comparison set

1,466Raw benchmark valueCI 1,460 - 1,472

Legal Research Bench

VALS-AI · Professional reasoning · Objective

Applied legal research tasks.

Rank #24 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 40.4%
Percentile: 67.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: legal_research; provider: OpenAI.

67.6% percentile inside its fair comparison set

40.4%Raw benchmark valueCI 33.7% - 47.1%

Harvey's Legal Agent Benchmark

VALS-AI · Professional reasoning · Objective

Completing legal work with documents, spreadsheets, presentations, and file-system tools.

Rank #34 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 3.8%
Percentile: 53.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: hlab; provider: OpenAI.

53.6% percentile inside its fair comparison set

3.8%Raw benchmark valueCI 1.4% - 6.1%

LegalBench

VALS-AI · Professional reasoning · Objective

Academic legal reasoning tasks.

Rank #12 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 86.5%
Percentile: 91.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: legal_bench; provider: OpenAI.

91.9% percentile inside its fair comparison set

86.5%Raw benchmark valueCI 85.7% - 87.3%

Finance Agent v2

VALS-AI · Professional reasoning · Objective

Core financial analyst tasks for agentic models.

Rank #30 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 51.8%
Percentile: 58%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: fabv2; provider: OpenAI.

58% percentile inside its fair comparison set

51.8%Raw benchmark valueCI 50.7% - 52.8%

MedCode

VALS-AI · Professional reasoning · Objective

Medical billing support and coding tasks.

Rank #26 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 49.1%
Percentile: 73.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: medcode; provider: OpenAI.

73.7% percentile inside its fair comparison set

49.1%Raw benchmark valueCI 44.8% - 53.4%

MedScribe

VALS-AI · Professional reasoning · Objective

Administrative documentation support for doctors.

Rank #18 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 86.9%
Percentile: 82.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: medscribe; provider: OpenAI.

82.3% percentile inside its fair comparison set

86.9%Raw benchmark valueCI 83.1% - 90.7%

SAGE

VALS-AI · Professional reasoning · Objective

Student Assessment with Generative Evaluation.

Rank #13 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 51.5%
Percentile: 85.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: sage; provider: OpenAI.

85.2% percentile inside its fair comparison set

51.5%Raw benchmark valueCI 43.8% - 59.3%

Public Benefits Bench

VALS-AI · Professional reasoning · Objective

Answering SNAP benefits questions across the public-benefits lifecycle.

Rank #28 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 60.9%
Percentile: 40%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: public-benefits-bench; provider: OpenAI.

40% percentile inside its fair comparison set

60.9%Raw benchmark valueCI 58.4% - 63.4%

SkillsBench

VALS-AI · Professional reasoning · Objective

Applied professional skills tasks.

Rank #4 · Source label: openai/gpt-5.5-codex

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 62.6%
Percentile: 90.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: skillsbench; provider: OpenAI.

90.9% percentile inside its fair comparison set

62.6%Raw benchmark valueCI 53.9% - 71.2%

TaxEval v2

VALS-AI · Professional reasoning · Objective

Answer quality on tax questions and responses.

Rank #25 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 75%
Percentile: 81.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: tax_eval_v2; provider: OpenAI.

81.4% percentile inside its fair comparison set

75%Raw benchmark valueCI 73.3% - 76.7%

Data analysis

LB · Professional reasoning · Objective

Structured data manipulation and table reasoning accuracy.

Rank #4 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 81.6%
Percentile: 95.4%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Category: Data Analysis. Tasks scored: 3.

95.4% percentile inside its fair comparison set

81.6%Raw benchmark value

Overall

LB · Professional reasoning · Objective

Average objective performance across LiveBench's current public category mix.

Rank #11 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 80.2%
Percentile: 84.6%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Category averages included: 7.

84.6% percentile inside its fair comparison set

80.2%Raw benchmark value

Consecutive events

LB · Professional reasoning · Objective

Objective consecutive events score in LiveBench.

Rank #21 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 88.8%
Percentile: 69.2%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: consecutive_events. Category: Data Analysis.

69.2% percentile inside its fair comparison set

88.8%Raw benchmark value

Table join

LB · Professional reasoning · Objective

Objective table join score in LiveBench.

Rank #5 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 55.9%
Percentile: 93.8%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: tablejoin. Category: Data Analysis.

93.8% percentile inside its fair comparison set

55.9%Raw benchmark value

Table reformat

LB · Professional reasoning · Objective

Objective table reformat score in LiveBench.

Rank #6 · Source label: gpt-5.5-xhigh

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 100%
Percentile: 100%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: tablereformat. Category: Data Analysis.

100% percentile inside its fair comparison set

100%Raw benchmark value

Public Benefits Bench v1

VALS-AI · Professional reasoning · Objective

Answering public-benefits questions across the benefits lifecycle.

Rank #8 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 57.2%
Percentile: 41.7%
Last updated: stale
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: public-benefits-bench-v1; provider: OpenAI.

41.7% percentile inside its fair comparison set

57.2%Raw benchmark valueCI 54.7% - 59.8%

Hallucination

BB · Professional reasoning · Rubric

Accuracy and fabrication resistance under prompts that invite unsupported claims.

Rank #13 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: BridgeBench
Raw value: 73.6%
Percentile: 64.7%
Last updated: archived
Eligibility: headline eligible

Parsed from the BridgeBench page for bridgebench-hallucination.

64.7% percentile inside its fair comparison set

73.6%Raw benchmark value

BS pushback

BB · Professional reasoning · Rubric

Resistance to confidently accepting bogus assumptions in expert-style prompts.

Rank #7 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: BridgeBench
Raw value: 88%
Percentile: 72.2%
Last updated: archived
Eligibility: headline eligible

Parsed from the BridgeBench page for bridgebench-pushback.

72.2% percentile inside its fair comparison set

88%Raw benchmark value

Search / tool use4 benchmarks82%

Tau2-Bench Telecom

AA · Search / tool use · Objective

It matters when the model must browse, call tools, and recover useful answers from external systems.

Rank #107 · Source label: GPT-5.5 (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 69.3%
Percentile: 68.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `tau2`.

68.7% percentile inside its fair comparison set

69.3%Raw benchmark value

Search Arena

AR · Search / tool use · Human

It matters when the model must browse, call tools, and recover useful answers from external systems.

Rank #3 · Source label: gpt-5.5-search

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,224
Percentile: 93.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-search`. Category: overall. Source rank: #3. Votes: 89873. Organization: openai. License: Proprietary.

93.9% percentile inside its fair comparison set

1,224Raw benchmark valueCI 1,218 - 1,229

Search Arena · No Style Control

AR · Search / tool use · Human

It matters when the model must browse, call tools, and recover useful answers from external systems.

Rank #3 · Source label: gpt-5.5-search

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,242
Percentile: 93.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-search`. Category: overall. Source rank: #3. Votes: 89873. Organization: openai. License: Proprietary.

93.9% percentile inside its fair comparison set

1,242Raw benchmark valueCI 1,237 - 1,247

BrowseComp

OFF · Search / tool use · Objective

It matters when the model must browse, call tools, and recover useful answers from external systems.

Rank #2

Official company resultmanual verifiedmanual verified

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Provider official evals
Raw value: 84.4%
Percentile: 83.3%
Last updated: archived
Eligibility: headline eligible

Verified against OpenAI's GPT-5.5 launch page dated April 23, 2026. Uses the published BrowseComp figure. Checked 2026-04-29. Verification: manual_public_page_verification.

83.3% percentile inside its fair comparison set

84.4%Raw benchmark value

Long context2 benchmarks77.3%

Long Context Reasoning

AA · Long context · Objective

It checks whether long-context claims survive contact with retrieval, memory, or long-document tasks.

Rank #176 · Source label: GPT-5.5 (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 64%
Percentile: 61.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `lcr`.

61.5% percentile inside its fair comparison set

64%Raw benchmark value

CorpFin v2

VALS-AI · Long context · Objective

It checks whether long-context claims survive contact with retrieval, memory, or long-document tasks.

Rank #9 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 68.4%
Percentile: 93.2%
Last updated: aging
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: corp_fin_v2; provider: OpenAI.

93.2% percentile inside its fair comparison set

68.4%Raw benchmark valueCI 66.6% - 70.2%

Vision understanding19 benchmarks85%

MMMU-Pro

AA · Vision understanding · Objective

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #86 · Source label: GPT-5.5 (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 71.4%
Percentile: 59.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `mmmuPro`.

59.1% percentile inside its fair comparison set

71.4%Raw benchmark value

Vision Arena

AR · Vision understanding · Human

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #16 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,287
Percentile: 89.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: overall. Source rank: #16. Votes: 27223. Organization: openai. License: Proprietary.

89.2% percentile inside its fair comparison set

1,287Raw benchmark valueCI 1,281 - 1,292

Vision Arena · Creative Writing Vision

AR · Vision understanding · Human

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #16 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,285
Percentile: 81.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: creative_writing_vision. Source rank: #16. Votes: 1882. Organization: openai. License: Proprietary.

81.7% percentile inside its fair comparison set

1,285Raw benchmark valueCI 1,270 - 1,299

Vision Arena · Diagram

AR · Vision understanding · Human

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #10 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,324
Percentile: 90.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: diagram. Source rank: #10. Votes: 7401. Organization: openai. License: Proprietary.

90.9% percentile inside its fair comparison set

1,324Raw benchmark valueCI 1,315 - 1,332

Vision Arena · English

AR · Vision understanding · Human

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #15 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,282
Percentile: 89.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: english. Source rank: #15. Votes: 10864. Organization: openai. License: Proprietary.

89.9% percentile inside its fair comparison set

1,282Raw benchmark valueCI 1,274 - 1,291

Vision Arena · Entity Recognition

AR · Vision understanding · Human

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #22 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,213
Percentile: 47.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: entity_recognition. Source rank: #24. Votes: 115. Organization: openai. License: Proprietary.

47.5% percentile inside its fair comparison set

1,213Raw benchmark valueCI 1,159 - 1,267

Vision Arena · Homework

AR · Vision understanding · Human

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #5 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,336
Percentile: 95.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: homework. Source rank: #5. Votes: 3312. Organization: openai. License: Proprietary.

95.8% percentile inside its fair comparison set

1,336Raw benchmark valueCI 1,325 - 1,347

Vision Arena · Humor

AR · Vision understanding · Human

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #12 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,285
Percentile: 84.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: humor. Source rank: #12. Votes: 993. Organization: openai. License: Proprietary.

84.9% percentile inside its fair comparison set

1,285Raw benchmark valueCI 1,266 - 1,305

Vision Arena · Ocr

AR · Vision understanding · Human

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #15 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,301
Percentile: 86%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: ocr. Source rank: #15. Votes: 19233. Organization: openai. License: Proprietary.

86% percentile inside its fair comparison set

1,301Raw benchmark valueCI 1,295 - 1,307

Vision Arena · No Style Control

AR · Vision understanding · Human

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #18 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,297
Percentile: 87.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: overall. Source rank: #19. Votes: 27223. Organization: openai. License: Proprietary.

87.8% percentile inside its fair comparison set

1,297Raw benchmark valueCI 1,291 - 1,303

Vision Arena · Creative Writing Vision · No Style Control

AR · Vision understanding · Human

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #14 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,305
Percentile: 84.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: creative_writing_vision. Source rank: #14. Votes: 1882. Organization: openai. License: Proprietary.

84.1% percentile inside its fair comparison set

1,305Raw benchmark valueCI 1,291 - 1,320

Vision Arena · Diagram · No Style Control

AR · Vision understanding · Human

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #12 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,326
Percentile: 88.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: diagram. Source rank: #12. Votes: 7401. Organization: openai. License: Proprietary.

88.9% percentile inside its fair comparison set

1,326Raw benchmark valueCI 1,317 - 1,334

Vision Arena · English · No Style Control

AR · Vision understanding · Human

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #20 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,293
Percentile: 86.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: english. Source rank: #21. Votes: 10864. Organization: openai. License: Proprietary.

86.2% percentile inside its fair comparison set

1,293Raw benchmark valueCI 1,285 - 1,301

Vision Arena · Entity Recognition · No Style Control

AR · Vision understanding · Human

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #26 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,225
Percentile: 37.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: entity_recognition. Source rank: #30. Votes: 115. Organization: openai. License: Proprietary.

37.5% percentile inside its fair comparison set

1,225Raw benchmark valueCI 1,171 - 1,278

Vision Arena · Homework · No Style Control

AR · Vision understanding · Human

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #14 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,334
Percentile: 86.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: homework. Source rank: #14. Votes: 3312. Organization: openai. License: Proprietary.

86.5% percentile inside its fair comparison set

1,334Raw benchmark valueCI 1,322 - 1,345

Vision Arena · Humor · No Style Control

AR · Vision understanding · Human

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #11 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,296
Percentile: 86.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: humor. Source rank: #12. Votes: 993. Organization: openai. License: Proprietary.

86.3% percentile inside its fair comparison set

1,296Raw benchmark valueCI 1,277 - 1,316

Vision Arena · Ocr · No Style Control

AR · Vision understanding · Human

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #17 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,308
Percentile: 84%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: ocr. Source rank: #18. Votes: 19233. Organization: openai. License: Proprietary.

84% percentile inside its fair comparison set

1,308Raw benchmark valueCI 1,302 - 1,314

MMMU Pro

VALS-AI · Vision understanding · Objective

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #9 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 88.3%
Percentile: 91.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: mmmu; provider: OpenAI.

91.6% percentile inside its fair comparison set

88.3%Raw benchmark valueCI 86.7% - 89.8%

MMMU-Pro

OFF · Vision understanding · Objective

It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.

Rank #1

Official company resultmanual verifiedmanual verified

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Provider official evals
Raw value: 81.2%
Percentile: 100%
Last updated: archived
Eligibility: headline eligible

Verified against OpenAI's GPT-5.5 launch page dated April 23, 2026. Uses the published no-tools MMMU Pro figure. Checked 2026-04-29. Verification: manual_public_page_verification.

100% percentile inside its fair comparison set

81.2%Raw benchmark value

Document understanding4 benchmarks80.7%

Document Arena

AR · Document understanding · Human

It matters when the job is reading PDFs, tables, forms, or mixed-layout documents rather than plain chat.

Rank #10 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,483
Percentile: 77.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: overall. Source rank: #10. Votes: 21279. Organization: openai. License: Proprietary.

77.5% percentile inside its fair comparison set

1,483Raw benchmark valueCI 1,476 - 1,489

Document Arena · No Style Control

AR · Document understanding · Human

It matters when the job is reading PDFs, tables, forms, or mixed-layout documents rather than plain chat.

Rank #8 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,486
Percentile: 82.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: overall. Source rank: #8. Votes: 21279. Organization: openai. License: Proprietary.

82.5% percentile inside its fair comparison set

1,486Raw benchmark valueCI 1,479 - 1,492

MortgageTax

VALS-AI · Document understanding · Objective

It matters when the job is reading PDFs, tables, forms, or mixed-layout documents rather than plain chat.

Rank #10 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 68.8%
Percentile: 89.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: mortgage_tax; provider: OpenAI.

89.5% percentile inside its fair comparison set

68.8%Raw benchmark valueCI 67% - 70.5%

Vals Multimodal Index

VALS-AI · Document understanding · Combined

It matters when the job is reading PDFs, tables, forms, or mixed-layout documents rather than plain chat.

Rank #9 · Source label: openai/gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 68
Percentile: 75%
Last updated: aging
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: vals_multimodal_index; provider: OpenAI.

75% percentile inside its fair comparison set

68Raw benchmark valueCI 65 - 71

Multilingual20 benchmarksn/a

Text Arena · Chinese

AR · Multilingual · Human

Observed user preference in Arena's Text Arena chinese leaderboard.

Rank #25 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,526
Percentile: 93.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: chinese. Source rank: #25. Votes: 4451. Organization: openai. License: Proprietary.

93.2% percentile inside its fair comparison set

1,526Raw benchmark valueCI 1,517 - 1,536

Text Arena · French

AR · Multilingual · Human

Observed user preference in Arena's Text Arena french leaderboard.

Rank #23 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,501
Percentile: 91.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: french. Source rank: #23. Votes: 2508. Organization: openai. License: Proprietary.

91.4% percentile inside its fair comparison set

1,501Raw benchmark valueCI 1,487 - 1,515

Text Arena · German

AR · Multilingual · Human

Observed user preference in Arena's Text Arena german leaderboard.

Rank #16 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,490
Percentile: 94.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: german. Source rank: #16. Votes: 1105. Organization: openai. License: Proprietary.

94.5% percentile inside its fair comparison set

1,490Raw benchmark valueCI 1,471 - 1,509

Text Arena · Japanese

AR · Multilingual · Human

Observed user preference in Arena's Text Arena japanese leaderboard.

Rank #6 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,503
Percentile: 97.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: japanese. Source rank: #6. Votes: 895. Organization: openai. License: Proprietary.

97.9% percentile inside its fair comparison set

1,503Raw benchmark valueCI 1,481 - 1,525

Text Arena · Korean

AR · Multilingual · Human

Observed user preference in Arena's Text Arena korean leaderboard.

Rank #25 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,445
Percentile: 90.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: korean. Source rank: #25. Votes: 1187. Organization: openai. License: Proprietary.

90.2% percentile inside its fair comparison set

1,445Raw benchmark valueCI 1,427 - 1,464

Text Arena · Non English

AR · Multilingual · Human

Observed user preference in Arena's Text Arena non_english leaderboard.

Rank #21 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,474
Percentile: 94.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: non_english. Source rank: #21. Votes: 37759. Organization: openai. License: Proprietary.

94.7% percentile inside its fair comparison set

1,474Raw benchmark valueCI 1,469 - 1,478

Text Arena · Polish

AR · Multilingual · Human

Observed user preference in Arena's Text Arena polish leaderboard.

Rank #29 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,479
Percentile: 85.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: polish. Source rank: #31. Votes: 1172. Organization: openai. License: Proprietary.

85.9% percentile inside its fair comparison set

1,479Raw benchmark valueCI 1,462 - 1,497

Text Arena · Russian

AR · Multilingual · Human

Observed user preference in Arena's Text Arena russian leaderboard.

Rank #31 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,480
Percentile: 91.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: russian. Source rank: #32. Votes: 7339. Organization: openai. License: Proprietary.

91.2% percentile inside its fair comparison set

1,480Raw benchmark valueCI 1,472 - 1,488

Text Arena · Spanish

AR · Multilingual · Human

Observed user preference in Arena's Text Arena spanish leaderboard.

Rank #29 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,469
Percentile: 89.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: spanish. Source rank: #29. Votes: 1937. Organization: openai. License: Proprietary.

89.2% percentile inside its fair comparison set

1,469Raw benchmark valueCI 1,455 - 1,484

Text Arena · Chinese · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena chinese leaderboard.

Rank #12 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,535
Percentile: 96.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: chinese. Source rank: #12. Votes: 4451. Organization: openai. License: Proprietary.

96.9% percentile inside its fair comparison set

1,535Raw benchmark valueCI 1,525 - 1,544

Text Arena · French · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena french leaderboard.

Rank #25 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,486
Percentile: 90.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: french. Source rank: #25. Votes: 2508. Organization: openai. License: Proprietary.

90.7% percentile inside its fair comparison set

1,486Raw benchmark valueCI 1,472 - 1,500

Text Arena · German · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena german leaderboard.

Rank #20 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,481
Percentile: 93%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: german. Source rank: #21. Votes: 1105. Organization: openai. License: Proprietary.

93% percentile inside its fair comparison set

1,481Raw benchmark valueCI 1,462 - 1,500

Text Arena · Japanese · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena japanese leaderboard.

Rank #9 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,497
Percentile: 96.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: japanese. Source rank: #9. Votes: 895. Organization: openai. License: Proprietary.

96.7% percentile inside its fair comparison set

1,497Raw benchmark valueCI 1,475 - 1,519

Text Arena · Korean · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena korean leaderboard.

Rank #12 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,461
Percentile: 95.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: korean. Source rank: #12. Votes: 1187. Organization: openai. License: Proprietary.

95.5% percentile inside its fair comparison set

1,461Raw benchmark valueCI 1,443 - 1,479

Text Arena · Non English · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena non_english leaderboard.

Rank #23 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,467
Percentile: 94.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: non_english. Source rank: #24. Votes: 37759. Organization: openai. License: Proprietary.

94.1% percentile inside its fair comparison set

1,467Raw benchmark valueCI 1,462 - 1,471

Text Arena · Polish · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena polish leaderboard.

Rank #29 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,465
Percentile: 85.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: polish. Source rank: #31. Votes: 1172. Organization: openai. License: Proprietary.

85.9% percentile inside its fair comparison set

1,465Raw benchmark valueCI 1,448 - 1,483

Text Arena · Russian · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena russian leaderboard.

Rank #27 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,473
Percentile: 92.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: russian. Source rank: #28. Votes: 7403. Organization: openai. License: Proprietary.

92.4% percentile inside its fair comparison set

1,473Raw benchmark valueCI 1,465 - 1,481

Text Arena · Spanish · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena spanish leaderboard.

Rank #30 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,468
Percentile: 88.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: spanish. Source rank: #31. Votes: 1937. Organization: openai. License: Proprietary.

88.8% percentile inside its fair comparison set

1,468Raw benchmark valueCI 1,453 - 1,482

Vision Arena · Chinese

AR · Multilingual · Human

Observed user preference in Arena's Vision Arena chinese leaderboard.

Rank #21 · Source label: gpt-5.5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,328
Percentile: 80.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5`. Category: chinese. Source rank: #22. Votes: 1711. Organization: openai. License: Proprietary.

80.8% percentile inside its fair comparison set

1,328Raw benchmark valueCI 1,311 - 1,345

Vision Arena · Chinese · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Vision Arena chinese leaderboard.

Rank #15 · Source label: gpt-5.5-high

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,359
Percentile: 86.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gpt-5.5-high`. Category: chinese. Source rank: #16. Votes: 1502. Organization: openai. License: Proprietary.

86.5% percentile inside its fair comparison set

1,359Raw benchmark valueCI 1,341 - 1,377

Source links and registry checks

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Arena

Oct 4, 2026

source →

official

Artificial Analysis

Oct 4, 2026

source →

official

LiveBench

Oct 4, 2026

source →

official

Scale Labs

Oct 4, 2026

source →

GPT-5.5

Current snapshot.

Source-linked scores by benchmark

Source links and registry checks

Loading model evidence.

GPT-5.5

Current snapshot.

Source-linked scores by benchmark

Source links and registry checks