Model profile · Google

Gemini 3 Flash

Closed weightspremium · registry tag 2026 current

Thin verified coverage

Reads as thin verified coverage across the resolved source data.

Visible coverage: 32.2%
Verified coverage: 32.2%
Spread: 93.8%
Last verified: Jun 20, 2026

63%bench fit

textcodevisiondocumentaudiosearch7 aliases47 official source links

Open compare

Data version

Current snapshot.

Data version Jun 20, 2026Model list checked9 providers · 1081 tracked modelsPage refreshed Jul 5, 2026

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

Source-linked scores by benchmark

Each row keeps the benchmark source, source type, raw metric, and percentile inside its fair comparison set.

Thin verified coverageThis model currently reads as thin verified coverage across the resolved source data.

Chat / text28 benchmarks80.6%

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 #68 · Source label: Gemini 3 Flash Preview (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 27
Percentile: 83%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `intelligenceIndex`.

83% percentile inside its fair comparison set

27Raw 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 #6 · Source label: Gemini 3 Flash Preview (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 45.5%
Percentile: 98.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `omniscienceAccuracy`.

98.3% percentile inside its fair comparison set

45.5%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 #244 · Source label: Gemini 3 Flash Preview (Reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 7.8%
Percentile: 18.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `omniscienceNonHallucination`.

18.5% percentile inside its fair comparison set

7.8%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 #71 · Source label: Gemini 3 Flash Preview (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 55.1%
Percentile: 77.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `ifbench`.

77.8% percentile inside its fair comparison set

55.1%Raw benchmark value

Blended 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 #177 · Source label: Gemini 3 Flash Preview (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: $1.1 /1M tokens
Percentile: 36.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `price1mBlended0To3To1`.

36.2% percentile inside its fair comparison set

$1.1 /1M tokensRaw 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 #161 · Source label: Gemini 3 Flash Preview (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: $0.5 /1M input tokens
Percentile: 42.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `price1mInputTokens`.

42.8% percentile inside its fair comparison set

$0.5 /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 #186 · Source label: Gemini 3 Flash Preview (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: $3 /1M output tokens
Percentile: 33%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `price1mOutputTokens`.

33% percentile inside its fair comparison set

$3 /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 #32 · Source label: Gemini 3 Flash Preview (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 186.8 tokens/s
Percentile: 85.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `medianOutputTokensPerSecond`.

85.2% percentile inside its fair comparison set

186.8 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 #175 · Source label: Gemini 3 Flash Preview (Reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 7.60s
Percentile: 17.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `medianTimeToFirstTokenSeconds`.

17.1% percentile inside its fair comparison set

7.60sRaw 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 #90 · Source label: Gemini 3 Flash Preview (Reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 7.60s
Percentile: 57.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `medianTimeToFirstAnswerTokenSeconds`.

57.6% percentile inside its fair comparison set

7.60sRaw 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 #15

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,473
Percentile: 95.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: overall. Source rank: #20. Votes: 30711. Organization: google. License: Proprietary.

95.7% percentile inside its fair comparison set

1,473Raw benchmark valueCI 1,469 - 1,478

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 #12

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,459
Percentile: 96.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: creative_writing. Source rank: #15. Votes: 4674. Organization: google. License: Proprietary.

96.6% percentile inside its fair comparison set

1,459Raw benchmark valueCI 1,449 - 1,468

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 #20

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,475
Percentile: 94.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: english. Source rank: #26. Votes: 14106. Organization: google. License: Proprietary.

94.2% percentile inside its fair comparison set

1,475Raw benchmark valueCI 1,469 - 1,481

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 #12

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,486
Percentile: 96.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: exclude_ties. Source rank: #14. Votes: 21818. Organization: google. License: Proprietary.

96.6% percentile inside its fair comparison set

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

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 #17

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,492
Percentile: 95.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: hard_prompts. Source rank: #22. Votes: 16572. Organization: google. License: Proprietary.

95.1% percentile inside its fair comparison set

1,492Raw benchmark valueCI 1,487 - 1,498

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 #24

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,488
Percentile: 92.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: hard_prompts_english. Source rank: #31. Votes: 7873. Organization: google. License: Proprietary.

92.9% percentile inside its fair comparison set

1,488Raw benchmark valueCI 1,480 - 1,495

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 #20

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,459
Percentile: 94.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: instruction_following. Source rank: #27. Votes: 8089. Organization: google. License: Proprietary.

94.2% percentile inside its fair comparison set

1,459Raw benchmark valueCI 1,451 - 1,466

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 #19

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,476
Percentile: 94.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: longer_query. Source rank: #26. Votes: 7873. Organization: google. License: Proprietary.

94.1% percentile inside its fair comparison set

1,476Raw benchmark valueCI 1,468 - 1,483

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 #15

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,484
Percentile: 95.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: multi_turn. Source rank: #18. Votes: 5309. Organization: google. License: Proprietary.

95.7% percentile inside its fair comparison set

1,484Raw benchmark valueCI 1,475 - 1,492

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 #14

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,466
Percentile: 96%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: overall. Source rank: #17. Votes: 30711. Organization: google. License: Proprietary.

96% percentile inside its fair comparison set

1,466Raw benchmark valueCI 1,462 - 1,471

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 #9

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,459
Percentile: 97.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: creative_writing. Source rank: #11. Votes: 4674. Organization: google. License: Proprietary.

97.5% percentile inside its fair comparison set

1,459Raw benchmark valueCI 1,450 - 1,468

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 #17

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,465
Percentile: 95.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: english. Source rank: #20. Votes: 14106. Organization: google. License: Proprietary.

95.1% percentile inside its fair comparison set

1,465Raw benchmark valueCI 1,459 - 1,471

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 #10

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,474
Percentile: 97.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: exclude_ties. Source rank: #13. Votes: 21818. Organization: google. License: Proprietary.

97.2% percentile inside its fair comparison set

1,474Raw benchmark valueCI 1,468 - 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

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,465
Percentile: 94.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: hard_prompts. Source rank: #26. Votes: 16572. Organization: google. License: Proprietary.

94.2% percentile inside its fair comparison set

1,465Raw benchmark valueCI 1,459 - 1,470

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 #29

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,458
Percentile: 91.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: hard_prompts_english. Source rank: #36. Votes: 7873. Organization: google. License: Proprietary.

91.4% percentile inside its fair comparison set

1,458Raw benchmark valueCI 1,451 - 1,465

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 #25

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,437
Percentile: 92.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: instruction_following. Source rank: #32. Votes: 8089. Organization: google. License: Proprietary.

92.6% percentile inside its fair comparison set

1,437Raw benchmark valueCI 1,430 - 1,445

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 #25

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,452
Percentile: 92.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: longer_query. Source rank: #32. Votes: 7873. Organization: google. License: Proprietary.

92.1% percentile inside its fair comparison set

1,452Raw benchmark valueCI 1,444 - 1,459

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 #13

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,473
Percentile: 96.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: multi_turn. Source rank: #16. Votes: 5309. Organization: google. License: Proprietary.

96.3% percentile inside its fair comparison set

1,473Raw benchmark valueCI 1,464 - 1,481

Coding16 benchmarks67.3%

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 #55 · Source label: Gemini 3 Flash Preview (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 31.8%
Percentile: 82.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `terminalbenchHard`.

82.5% percentile inside its fair comparison set

31.8%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 #10 · Source label: Gemini 3 Flash Preview (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 49.9%
Percentile: 97.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `scicode`.

97.8% percentile inside its fair comparison set

49.9%Raw benchmark value

Code 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 #24

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,437
Percentile: 68.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: overall. Source rank: #31. Votes: 13277. Organization: google. License: Proprietary.

68.5% percentile inside its fair comparison set

1,437Raw benchmark valueCI 1,430 - 1,445

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 #24

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,437
Percentile: 68.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: webdev. Source rank: #31. Votes: 13277. Organization: google. License: Proprietary.

68.5% percentile inside its fair comparison set

1,437Raw benchmark valueCI 1,430 - 1,445

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 #25

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,440
Percentile: 67.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: webdev-html. Source rank: #32. Votes: 9253. Organization: google. License: Proprietary.

67.1% percentile inside its fair comparison set

1,440Raw benchmark valueCI 1,430 - 1,450

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 #23

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,431
Percentile: 62.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: webdev-react. Source rank: #29. Votes: 3993. Organization: google. License: Proprietary.

62.7% percentile inside its fair comparison set

1,431Raw benchmark valueCI 1,420 - 1,441

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 #22 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 6.4%
Percentile: 4.5%
Last updated: recent
Eligibility: headline eligible

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

4.5% percentile inside its fair comparison set

6.4%Raw benchmark valueCI 4.2% - 8.5%

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 #16 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 85.6%
Percentile: 83.3%
Last updated: recent
Eligibility: headline eligible

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

83.3% percentile inside its fair comparison set

85.6%Raw benchmark valueCI 83.6% - 87.6%

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 #17 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 0%
Percentile: 70%
Last updated: recent
Eligibility: headline eligible

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

70% percentile inside its fair comparison set

0%Raw benchmark valueCI 0% - 0%

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 #19 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 75%
Percentile: 66.7%
Last updated: recent
Eligibility: headline eligible

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

66.7% percentile inside its fair comparison set

75%Raw benchmark valueCI 71.2% - 78.8%

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 #15 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 53.9%
Percentile: 48.1%
Last updated: recent
Eligibility: headline eligible

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

48.1% percentile inside its fair comparison set

53.9%Raw benchmark valueCI 51.4% - 56.5%

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 #29 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 20.2%
Percentile: 42.9%
Last updated: recent
Eligibility: headline eligible

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

42.9% percentile inside its fair comparison set

20.2%Raw benchmark valueCI 12.5% - 27.9%

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 #26

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,509
Percentile: 92.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: coding. Source rank: #33. Votes: 6387. Organization: google. License: Proprietary.

92.2% percentile inside its fair comparison set

1,509Raw benchmark valueCI 1,501 - 1,517

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 #36

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,463
Percentile: 89.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: coding. Source rank: #44. Votes: 6387. Organization: google. License: Proprietary.

89.1% percentile inside its fair comparison set

1,463Raw benchmark valueCI 1,455 - 1,471

IOI

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 #7 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 39.1%
Percentile: 86.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: ioi; provider: Google.

86.4% percentile inside its fair comparison set

39.1%Raw benchmark valueCI 18.4% - 59.8%

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 #10

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,447
Percentile: 47.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: image_to_webdev. Source rank: #13. Votes: 4435. Organization: google. License: Proprietary.

47.1% percentile inside its fair comparison set

1,447Raw benchmark valueCI 1,436 - 1,457

Reasoning / math / science8 benchmarks82%

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 #61 · Source label: Gemini 3 Flash Preview (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 14.1%
Percentile: 83.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `hle`.

83.8% percentile inside its fair comparison set

14.1%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 #57 · Source label: Gemini 3 Flash Preview (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 81.2%
Percentile: 85%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `gpqa`.

85% percentile inside its fair comparison set

81.2%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 #39 · Source label: Gemini 3 Flash Preview (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 1.4%
Percentile: 88.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `critpt`.

88.1% percentile inside its fair comparison set

1.4%Raw benchmark value

ProofBench

VALS-AI · Reasoning / math / science · Objective

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

Rank #23 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 15%
Percentile: 37.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: proof_bench; provider: Google.

37.1% percentile inside its fair comparison set

15%Raw benchmark valueCI 8% - 22%

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 #19 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 87.9%
Percentile: 79.8%
Last updated: recent
Eligibility: headline eligible

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

79.8% percentile inside its fair comparison set

87.9%Raw benchmark valueCI 84.7% - 91.1%

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 #9 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 88.6%
Percentile: 91%
Last updated: recent
Eligibility: headline eligible

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

91% percentile inside its fair comparison set

88.6%Raw benchmark valueCI 88% - 89.2%

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 #16

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,476
Percentile: 95.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: math. Source rank: #20. Votes: 2002. Organization: google. License: Proprietary.

95.2% percentile inside its fair comparison set

1,476Raw benchmark valueCI 1,463 - 1,489

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 #14

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,474
Percentile: 95.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: math. Source rank: #18. Votes: 2002. Organization: google. License: Proprietary.

95.9% percentile inside its fair comparison set

1,474Raw benchmark valueCI 1,460 - 1,487

Professional reasoning27 benchmarks84.5%

APEX-Agents-AA

AA · Professional reasoning · Objective

Long-horizon agentic task completion.

Rank #9 · Source label: Gemini 3 Flash Preview (Reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 27.7%
Percentile: 66.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `apexAgents`.

66.7% percentile inside its fair comparison set

27.7%Raw benchmark value

Vals Index

VALS-AI · Professional reasoning · Combined

Weighted model performance across economically relevant Vals tasks.

Rank #18 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 50
Percentile: 34.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: vals_index; provider: Google.

34.6% percentile inside its fair comparison set

50Raw benchmark valueCI 47 - 53

LegalBench

VALS-AI · Professional reasoning · Objective

Academic legal reasoning tasks.

Rank #4 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 86.9%
Percentile: 96.7%
Last updated: recent
Eligibility: headline eligible

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

96.7% percentile inside its fair comparison set

86.9%Raw benchmark valueCI 86.2% - 87.6%

Finance Agent v2

VALS-AI · Professional reasoning · Objective

Core financial analyst tasks for agentic models.

Rank #14 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 42.6%
Percentile: 48%
Last updated: recent
Eligibility: headline eligible

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

48% percentile inside its fair comparison set

42.6%Raw benchmark valueCI 41.7% - 43.4%

TaxEval v2

VALS-AI · Professional reasoning · Objective

Answer quality on tax questions and responses.

Rank #28 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 73.9%
Percentile: 70.3%
Last updated: recent
Eligibility: headline eligible

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

70.3% percentile inside its fair comparison set

73.9%Raw benchmark valueCI 72.2% - 75.6%

MedCode

VALS-AI · Professional reasoning · Objective

Medical billing support and coding tasks.

Rank #3 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 55.9%
Percentile: 96.1%
Last updated: recent
Eligibility: headline eligible

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

96.1% percentile inside its fair comparison set

55.9%Raw benchmark valueCI 51.8% - 60.1%

MedScribe

VALS-AI · Professional reasoning · Objective

Administrative documentation support for doctors.

Rank #45 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 69.9%
Percentile: 12%
Last updated: recent
Eligibility: headline eligible

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

12% percentile inside its fair comparison set

69.9%Raw benchmark valueCI 66.2% - 73.6%

Text Arena · Expert

AR · Professional reasoning · Human

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

Rank #19

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,497
Percentile: 93.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: expert. Source rank: #25. Votes: 1932. Organization: google. License: Proprietary.

93.5% percentile inside its fair comparison set

1,497Raw benchmark valueCI 1,483 - 1,510

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 #17

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,470
Percentile: 95%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: industry_business_and_management_and_financial_operations. Source rank: #21. Votes: 5623. Organization: google. License: Proprietary.

95% percentile inside its fair comparison set

1,470Raw benchmark valueCI 1,462 - 1,479

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 #11

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,454
Percentile: 96.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: industry_entertainment_and_sports_and_media. Source rank: #13. Votes: 5585. Organization: google. License: Proprietary.

96.9% percentile inside its fair comparison set

1,454Raw benchmark valueCI 1,445 - 1,462

Text Arena · Industry Legal And Government

AR · Professional reasoning · Human

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

Rank #10

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,487
Percentile: 97%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: industry_legal_and_government. Source rank: #12. Votes: 2310. Organization: google. License: Proprietary.

97% percentile inside its fair comparison set

1,487Raw benchmark valueCI 1,475 - 1,500

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 #21

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,486
Percentile: 93.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: industry_life_and_physical_and_social_science. Source rank: #25. Votes: 4888. Organization: google. License: Proprietary.

93.8% percentile inside its fair comparison set

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

Text Arena · Industry Mathematical

AR · Professional reasoning · Human

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

Rank #24

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,472
Percentile: 92.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: industry_mathematical. Source rank: #28. Votes: 1414. Organization: google. License: Proprietary.

92.5% percentile inside its fair comparison set

1,472Raw benchmark valueCI 1,456 - 1,488

Text Arena · Industry Medicine And Healthcare

AR · Professional reasoning · Human

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

Rank #23

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,484
Percentile: 92.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: industry_medicine_and_healthcare. Source rank: #29. Votes: 2069. Organization: google. License: Proprietary.

92.5% percentile inside its fair comparison set

1,484Raw benchmark valueCI 1,471 - 1,498

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 #21

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,501
Percentile: 93.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: industry_software_and_it_services. Source rank: #27. Votes: 10515. Organization: google. License: Proprietary.

93.8% percentile inside its fair comparison set

1,501Raw benchmark valueCI 1,494 - 1,508

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 #14

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,461
Percentile: 96%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: industry_writing_and_literature_and_language. Source rank: #18. Votes: 6818. Organization: google. License: Proprietary.

96% percentile inside its fair comparison set

1,461Raw benchmark valueCI 1,453 - 1,469

Text Arena · Expert · No Style Control

AR · Professional reasoning · Human

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

Rank #28

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,463
Percentile: 90.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: expert. Source rank: #36. Votes: 1932. Organization: google. License: Proprietary.

90.2% percentile inside its fair comparison set

1,463Raw benchmark valueCI 1,450 - 1,477

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 #17

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,450
Percentile: 95%
Last updated: recent
Eligibility: headline eligible

95% percentile inside its fair comparison set

1,450Raw benchmark valueCI 1,442 - 1,458

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 #9

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,449
Percentile: 97.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: industry_entertainment_and_sports_and_media. Source rank: #11. Votes: 5585. Organization: google. License: Proprietary.

97.5% percentile inside its fair comparison set

1,449Raw benchmark valueCI 1,440 - 1,457

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 #13

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,477
Percentile: 96%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: industry_legal_and_government. Source rank: #15. Votes: 2310. Organization: google. License: Proprietary.

96% percentile inside its fair comparison set

1,477Raw benchmark valueCI 1,465 - 1,490

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 #18

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,472
Percentile: 94.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: industry_life_and_physical_and_social_science. Source rank: #22. Votes: 4888. Organization: google. License: Proprietary.

94.7% percentile inside its fair comparison set

1,472Raw benchmark valueCI 1,464 - 1,481

Text Arena · Industry Mathematical · No Style Control

AR · Professional reasoning · Human

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

Rank #22

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,465
Percentile: 93.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: industry_mathematical. Source rank: #26. Votes: 1414. Organization: google. License: Proprietary.

93.2% percentile inside its fair comparison set

1,465Raw benchmark valueCI 1,450 - 1,481

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 #22

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,465
Percentile: 92.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: industry_medicine_and_healthcare. Source rank: #25. Votes: 2069. Organization: google. License: Proprietary.

92.9% percentile inside its fair comparison set

1,465Raw benchmark valueCI 1,452 - 1,478

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 #24

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,471
Percentile: 92.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: industry_software_and_it_services. Source rank: #30. Votes: 10515. Organization: google. License: Proprietary.

92.9% percentile inside its fair comparison set

1,471Raw benchmark valueCI 1,465 - 1,478

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 #13

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,459
Percentile: 96.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: industry_writing_and_literature_and_language. Source rank: #16. Votes: 6818. Organization: google. License: Proprietary.

96.3% percentile inside its fair comparison set

1,459Raw benchmark valueCI 1,451 - 1,467

SAGE

VALS-AI · Professional reasoning · Objective

Student Assessment with Generative Evaluation.

Rank #6 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 51.8%
Percentile: 88.9%
Last updated: recent
Eligibility: headline eligible

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

88.9% percentile inside its fair comparison set

51.8%Raw benchmark valueCI 45.2% - 58.5%

Poker Agent

VALS-AI · Professional reasoning · Objective

Agent profit in poker-style strategic play.

Rank #6 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 1,100.2 score
Percentile: 68.8%
Last updated: archived
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: poker_agent; provider: unknown.

68.8% percentile inside its fair comparison set

1,100.2 scoreRaw benchmark valueCI 1,100.2 score - 1,100.2 score

Search / tool use3 benchmarks62.1%

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 #146 · Source label: Gemini 3 Flash Preview (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 43.3%
Percentile: 53.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `tau2`.

53.1% percentile inside its fair comparison set

43.3%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 #10 · Source label: gemini-3-flash-grounding

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,194
Percentile: 70%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash-grounding`. Category: overall. Source rank: #11. Votes: 93980. Organization: google. License: Proprietary.

70% percentile inside its fair comparison set

1,194Raw benchmark valueCI 1,189 - 1,199

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 #12 · Source label: gemini-3-flash-grounding

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,201
Percentile: 63.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash-grounding`. Category: overall. Source rank: #12. Votes: 93980. Organization: google. License: Proprietary.

63.3% percentile inside its fair comparison set

1,201Raw benchmark valueCI 1,196 - 1,206

Long context2 benchmarks78.4%

Long Context Reasoning

AA · Long context · Objective

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

Rank #98 · Source label: Gemini 3 Flash Preview (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 48%
Percentile: 69.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `lcr`.

69.2% percentile inside its fair comparison set

48%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 #12 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 66.4%
Percentile: 87.5%
Last updated: recent
Eligibility: headline eligible

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

87.5% percentile inside its fair comparison set

66.4%Raw benchmark valueCI 64.6% - 68.3%

Vision understanding22 benchmarks88.9%

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 #7 · Source label: Gemini 3 Flash Preview (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 78.6%
Percentile: 95.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `mmmuPro`.

95.6% percentile inside its fair comparison set

78.6%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 #14

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,272
Percentile: 88.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: overall. Source rank: #18. Votes: 28104. Organization: google. License: Proprietary.

88.1% percentile inside its fair comparison set

1,272Raw benchmark valueCI 1,266 - 1,278

Vision Arena · Captioning

AR · Vision understanding · Human

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

Rank #6

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,235
Percentile: 80.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: captioning. Source rank: #6. Votes: 163. Organization: google. License: Proprietary.

80.8% percentile inside its fair comparison set

1,235Raw benchmark valueCI 1,192 - 1,277

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 #6 · Source label: gemini-3-flash (thinking-minimal)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,283
Percentile: 90.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash (thinking-minimal)`. Category: creative_writing_vision. Source rank: #8. Votes: 1624. Organization: google. License: Proprietary.

90.9% percentile inside its fair comparison set

1,283Raw benchmark valueCI 1,268 - 1,298

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 #14

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,296
Percentile: 81.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: diagram. Source rank: #19. Votes: 7012. Organization: google. License: Proprietary.

81.4% percentile inside its fair comparison set

1,296Raw benchmark valueCI 1,288 - 1,305

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 #16

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,261
Percentile: 86.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: english. Source rank: #20. Votes: 11898. Organization: google. License: Proprietary.

86.2% percentile inside its fair comparison set

1,261Raw benchmark valueCI 1,253 - 1,269

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 #2

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,296
Percentile: 96.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: entity_recognition. Source rank: #2. Votes: 233. Organization: google. License: Proprietary.

96.9% percentile inside its fair comparison set

1,296Raw benchmark valueCI 1,260 - 1,332

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 #7

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,312
Percentile: 91.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: homework. Source rank: #10. Votes: 4223. Organization: google. License: Proprietary.

91.2% percentile inside its fair comparison set

1,312Raw benchmark valueCI 1,302 - 1,322

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 #9

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,268
Percentile: 83.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: humor. Source rank: #12. Votes: 958. Organization: google. License: Proprietary.

83.7% percentile inside its fair comparison set

1,268Raw benchmark valueCI 1,248 - 1,287

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 #13

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,286
Percentile: 82.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: ocr. Source rank: #17. Votes: 19274. Organization: google. License: Proprietary.

82.9% percentile inside its fair comparison set

1,286Raw benchmark valueCI 1,280 - 1,291

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 #11

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,285
Percentile: 90.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: overall. Source rank: #15. Votes: 28104. Organization: google. License: Proprietary.

90.8% percentile inside its fair comparison set

1,285Raw benchmark valueCI 1,279 - 1,290

Vision Arena · Captioning · 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 #5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,282
Percentile: 84.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: captioning. Source rank: #5. Votes: 163. Organization: google. License: Proprietary.

84.6% percentile inside its fair comparison set

1,282Raw benchmark valueCI 1,240 - 1,323

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 #6

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,297
Percentile: 90.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: creative_writing_vision. Source rank: #8. Votes: 1817. Organization: google. License: Proprietary.

90.9% percentile inside its fair comparison set

1,297Raw benchmark valueCI 1,283 - 1,311

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

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,294
Percentile: 84.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: diagram. Source rank: #17. Votes: 7012. Organization: google. License: Proprietary.

84.3% percentile inside its fair comparison set

1,294Raw benchmark valueCI 1,286 - 1,302

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 #15

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,274
Percentile: 87.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: english. Source rank: #19. Votes: 11898. Organization: google. License: Proprietary.

87.2% percentile inside its fair comparison set

1,274Raw benchmark valueCI 1,266 - 1,282

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 #2

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,338
Percentile: 96.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: entity_recognition. Source rank: #2. Votes: 233. Organization: google. License: Proprietary.

96.9% percentile inside its fair comparison set

1,338Raw benchmark valueCI 1,302 - 1,375

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 #9

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,317
Percentile: 88.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: homework. Source rank: #12. Votes: 4223. Organization: google. License: Proprietary.

88.2% percentile inside its fair comparison set

1,317Raw benchmark valueCI 1,307 - 1,327

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 #6

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,289
Percentile: 89.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: humor. Source rank: #8. Votes: 958. Organization: google. License: Proprietary.

89.8% percentile inside its fair comparison set

1,289Raw benchmark valueCI 1,270 - 1,309

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 #12

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,291
Percentile: 84.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: ocr. Source rank: #16. Votes: 19274. Organization: google. License: Proprietary.

84.3% percentile inside its fair comparison set

1,291Raw benchmark valueCI 1,286 - 1,297

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 #5 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 87.6%
Percentile: 93.1%
Last updated: recent
Eligibility: headline eligible

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

93.1% percentile inside its fair comparison set

87.6%Raw benchmark valueCI 86.1% - 89.2%

Vision Arena · Creative Writing

AR · Vision understanding · Human

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

Rank #3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,289
Percentile: 93.8%
Last updated: archived
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: creative_writing. Source rank: #3. Votes: 217. Organization: google. License: Proprietary.

93.8% percentile inside its fair comparison set

1,289Raw benchmark valueCI 1,249 - 1,329

Vision Arena · Creative Writing · 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 #3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,306
Percentile: 93.8%
Last updated: archived
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: creative_writing. Source rank: #3. Votes: 217. Organization: google. License: Proprietary.

93.8% percentile inside its fair comparison set

1,306Raw benchmark valueCI 1,267 - 1,346

Document understanding4 benchmarks43.8%

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 #17

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,442
Percentile: 33.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: overall. Source rank: #20. Votes: 7188. Organization: google. License: Proprietary.

33.3% percentile inside its fair comparison set

1,442Raw benchmark valueCI 1,433 - 1,451

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 #21

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,413
Percentile: 16.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: overall. Source rank: #24. Votes: 7188. Organization: google. License: Proprietary.

16.7% percentile inside its fair comparison set

1,413Raw benchmark valueCI 1,404 - 1,422

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 #13 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 52
Percentile: 36.8%
Last updated: recent
Eligibility: headline eligible

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

36.8% percentile inside its fair comparison set

52Raw benchmark valueCI 49 - 55

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 #8 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 68.7%
Percentile: 88.3%
Last updated: recent
Eligibility: headline eligible

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

88.3% percentile inside its fair comparison set

68.7%Raw benchmark valueCI 66.9% - 70.5%

Multilingual16 benchmarks95.4%

Text Arena · Chinese

AR · Multilingual · Human

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

Rank #11

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,521
Percentile: 96.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: chinese. Source rank: #15. Votes: 1513. Organization: google. License: Proprietary.

96.6% percentile inside its fair comparison set

1,521Raw benchmark valueCI 1,506 - 1,537

Text Arena · French

AR · Multilingual · Human

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

Rank #8

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,506
Percentile: 96.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: french. Source rank: #9. Votes: 754. Organization: google. License: Proprietary.

96.8% percentile inside its fair comparison set

1,506Raw benchmark valueCI 1,483 - 1,529

Text Arena · German

AR · Multilingual · Human

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

Rank #6

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,494
Percentile: 97.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: german. Source rank: #7. Votes: 531. Organization: google. License: Proprietary.

97.9% percentile inside its fair comparison set

1,494Raw benchmark valueCI 1,468 - 1,520

Text Arena · Japanese

AR · Multilingual · Human

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

Rank #6

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,492
Percentile: 97.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: japanese. Source rank: #6. Votes: 272. Organization: google. License: Proprietary.

97.5% percentile inside its fair comparison set

1,492Raw benchmark valueCI 1,454 - 1,530

Text Arena · Korean

AR · Multilingual · Human

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

Rank #12

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,447
Percentile: 94.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: korean. Source rank: #13. Votes: 537. Organization: google. License: Proprietary.

94.7% percentile inside its fair comparison set

1,447Raw benchmark valueCI 1,421 - 1,474

Text Arena · Russian

AR · Multilingual · Human

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

Rank #10

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,486
Percentile: 96.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: russian. Source rank: #13. Votes: 3264. Organization: google. License: Proprietary.

96.9% percentile inside its fair comparison set

1,486Raw benchmark valueCI 1,475 - 1,496

Text Arena · Spanish

AR · Multilingual · Human

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

Rank #15

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,468
Percentile: 93.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: spanish. Source rank: #17. Votes: 832. Organization: google. License: Proprietary.

93.5% percentile inside its fair comparison set

1,468Raw benchmark valueCI 1,447 - 1,489

Text Arena · Chinese · No Style Control

AR · Multilingual · Human

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

Rank #12

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,514
Percentile: 96.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: chinese. Source rank: #16. Votes: 1513. Organization: google. License: Proprietary.

96.3% percentile inside its fair comparison set

1,514Raw benchmark valueCI 1,498 - 1,529

Text Arena · French · No Style Control

AR · Multilingual · Human

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

Rank #10

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,493
Percentile: 95.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: french. Source rank: #12. Votes: 754. Organization: google. License: Proprietary.

95.8% percentile inside its fair comparison set

1,493Raw benchmark valueCI 1,470 - 1,516

Text Arena · German · No Style Control

AR · Multilingual · Human

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

Rank #5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,493
Percentile: 98.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: german. Source rank: #5. Votes: 531. Organization: google. License: Proprietary.

98.3% percentile inside its fair comparison set

1,493Raw benchmark valueCI 1,467 - 1,519

Text Arena · Japanese · No Style Control

AR · Multilingual · Human

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

Rank #5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,495
Percentile: 98%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: japanese. Source rank: #5. Votes: 272. Organization: google. License: Proprietary.

98% percentile inside its fair comparison set

1,495Raw benchmark valueCI 1,456 - 1,533

Text Arena · Korean · No Style Control

AR · Multilingual · Human

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

Rank #9

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,446
Percentile: 96.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: korean. Source rank: #9. Votes: 537. Organization: google. License: Proprietary.

96.2% percentile inside its fair comparison set

1,446Raw benchmark valueCI 1,419 - 1,473

Text Arena · Russian · No Style Control

AR · Multilingual · Human

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

Rank #10

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,479
Percentile: 96.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: russian. Source rank: #13. Votes: 3264. Organization: google. License: Proprietary.

96.9% percentile inside its fair comparison set

1,479Raw benchmark valueCI 1,468 - 1,489

Text Arena · Spanish · No Style Control

AR · Multilingual · Human

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

Rank #17

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,460
Percentile: 92.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: spanish. Source rank: #19. Votes: 832. Organization: google. License: Proprietary.

92.5% percentile inside its fair comparison set

1,460Raw benchmark valueCI 1,439 - 1,481

Vision Arena · Chinese

AR · Multilingual · Human

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

Rank #8

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,323
Percentile: 90.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: chinese. Source rank: #10. Votes: 1512. Organization: google. License: Proprietary.

90.9% percentile inside its fair comparison set

1,323Raw benchmark valueCI 1,303 - 1,343

Vision Arena · Chinese · No Style Control

AR · Multilingual · Human

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

Rank #11

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,337
Percentile: 87%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: chinese. Source rank: #15. Votes: 1512. Organization: google. License: Proprietary.

87% percentile inside its fair comparison set

1,337Raw benchmark valueCI 1,317 - 1,356

Source links and registry checks

official

Google Gemini models docs

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Arena

Jun 20, 2026

source →

official

Artificial Analysis

Jun 20, 2026

source →

Model profile · Google

Gemini 3 Flash

Closed weightspremium · registry tag 2026 current

Thin verified coverage

Reads as thin verified coverage across the resolved source data.

Visible coverage: 32.2%
Verified coverage: 32.2%
Spread: 93.8%
Last verified: Jun 20, 2026

63%bench fit

textcodevisiondocumentaudiosearch7 aliases47 official source links

Open compare

Data version

Current snapshot.

Data version Jun 20, 2026Model list checked9 providers · 1081 tracked modelsPage refreshed Jul 5, 2026

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

Source-linked scores by benchmark

Each row keeps the benchmark source, source type, raw metric, and percentile inside its fair comparison set.

Thin verified coverageThis model currently reads as thin verified coverage across the resolved source data.

Chat / text28 benchmarks80.6%

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 #68 · Source label: Gemini 3 Flash Preview (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 27
Percentile: 83%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `intelligenceIndex`.

83% percentile inside its fair comparison set

27Raw 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 #6 · Source label: Gemini 3 Flash Preview (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 45.5%
Percentile: 98.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `omniscienceAccuracy`.

98.3% percentile inside its fair comparison set

45.5%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 #244 · Source label: Gemini 3 Flash Preview (Reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 7.8%
Percentile: 18.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `omniscienceNonHallucination`.

18.5% percentile inside its fair comparison set

7.8%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 #71 · Source label: Gemini 3 Flash Preview (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 55.1%
Percentile: 77.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `ifbench`.

77.8% percentile inside its fair comparison set

55.1%Raw benchmark value

Blended 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 #177 · Source label: Gemini 3 Flash Preview (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: $1.1 /1M tokens
Percentile: 36.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `price1mBlended0To3To1`.

36.2% percentile inside its fair comparison set

$1.1 /1M tokensRaw 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 #161 · Source label: Gemini 3 Flash Preview (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: $0.5 /1M input tokens
Percentile: 42.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `price1mInputTokens`.

42.8% percentile inside its fair comparison set

$0.5 /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 #186 · Source label: Gemini 3 Flash Preview (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: $3 /1M output tokens
Percentile: 33%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `price1mOutputTokens`.

33% percentile inside its fair comparison set

$3 /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 #32 · Source label: Gemini 3 Flash Preview (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 186.8 tokens/s
Percentile: 85.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `medianOutputTokensPerSecond`.

85.2% percentile inside its fair comparison set

186.8 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 #175 · Source label: Gemini 3 Flash Preview (Reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 7.60s
Percentile: 17.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `medianTimeToFirstTokenSeconds`.

17.1% percentile inside its fair comparison set

7.60sRaw 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 #90 · Source label: Gemini 3 Flash Preview (Reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 7.60s
Percentile: 57.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `medianTimeToFirstAnswerTokenSeconds`.

57.6% percentile inside its fair comparison set

7.60sRaw 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 #15

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,473
Percentile: 95.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: overall. Source rank: #20. Votes: 30711. Organization: google. License: Proprietary.

95.7% percentile inside its fair comparison set

1,473Raw benchmark valueCI 1,469 - 1,478

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 #12

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,459
Percentile: 96.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: creative_writing. Source rank: #15. Votes: 4674. Organization: google. License: Proprietary.

96.6% percentile inside its fair comparison set

1,459Raw benchmark valueCI 1,449 - 1,468

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 #20

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,475
Percentile: 94.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: english. Source rank: #26. Votes: 14106. Organization: google. License: Proprietary.

94.2% percentile inside its fair comparison set

1,475Raw benchmark valueCI 1,469 - 1,481

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 #12

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,486
Percentile: 96.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: exclude_ties. Source rank: #14. Votes: 21818. Organization: google. License: Proprietary.

96.6% percentile inside its fair comparison set

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

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 #17

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,492
Percentile: 95.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: hard_prompts. Source rank: #22. Votes: 16572. Organization: google. License: Proprietary.

95.1% percentile inside its fair comparison set

1,492Raw benchmark valueCI 1,487 - 1,498

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 #24

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,488
Percentile: 92.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: hard_prompts_english. Source rank: #31. Votes: 7873. Organization: google. License: Proprietary.

92.9% percentile inside its fair comparison set

1,488Raw benchmark valueCI 1,480 - 1,495

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 #20

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,459
Percentile: 94.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: instruction_following. Source rank: #27. Votes: 8089. Organization: google. License: Proprietary.

94.2% percentile inside its fair comparison set

1,459Raw benchmark valueCI 1,451 - 1,466

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 #19

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,476
Percentile: 94.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: longer_query. Source rank: #26. Votes: 7873. Organization: google. License: Proprietary.

94.1% percentile inside its fair comparison set

1,476Raw benchmark valueCI 1,468 - 1,483

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 #15

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,484
Percentile: 95.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: multi_turn. Source rank: #18. Votes: 5309. Organization: google. License: Proprietary.

95.7% percentile inside its fair comparison set

1,484Raw benchmark valueCI 1,475 - 1,492

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 #14

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,466
Percentile: 96%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: overall. Source rank: #17. Votes: 30711. Organization: google. License: Proprietary.

96% percentile inside its fair comparison set

1,466Raw benchmark valueCI 1,462 - 1,471

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 #9

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,459
Percentile: 97.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: creative_writing. Source rank: #11. Votes: 4674. Organization: google. License: Proprietary.

97.5% percentile inside its fair comparison set

1,459Raw benchmark valueCI 1,450 - 1,468

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 #17

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,465
Percentile: 95.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: english. Source rank: #20. Votes: 14106. Organization: google. License: Proprietary.

95.1% percentile inside its fair comparison set

1,465Raw benchmark valueCI 1,459 - 1,471

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 #10

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,474
Percentile: 97.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: exclude_ties. Source rank: #13. Votes: 21818. Organization: google. License: Proprietary.

97.2% percentile inside its fair comparison set

1,474Raw benchmark valueCI 1,468 - 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

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,465
Percentile: 94.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: hard_prompts. Source rank: #26. Votes: 16572. Organization: google. License: Proprietary.

94.2% percentile inside its fair comparison set

1,465Raw benchmark valueCI 1,459 - 1,470

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 #29

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,458
Percentile: 91.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: hard_prompts_english. Source rank: #36. Votes: 7873. Organization: google. License: Proprietary.

91.4% percentile inside its fair comparison set

1,458Raw benchmark valueCI 1,451 - 1,465

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 #25

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,437
Percentile: 92.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: instruction_following. Source rank: #32. Votes: 8089. Organization: google. License: Proprietary.

92.6% percentile inside its fair comparison set

1,437Raw benchmark valueCI 1,430 - 1,445

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 #25

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,452
Percentile: 92.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: longer_query. Source rank: #32. Votes: 7873. Organization: google. License: Proprietary.

92.1% percentile inside its fair comparison set

1,452Raw benchmark valueCI 1,444 - 1,459

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 #13

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,473
Percentile: 96.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: multi_turn. Source rank: #16. Votes: 5309. Organization: google. License: Proprietary.

96.3% percentile inside its fair comparison set

1,473Raw benchmark valueCI 1,464 - 1,481

Coding16 benchmarks67.3%

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 #55 · Source label: Gemini 3 Flash Preview (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 31.8%
Percentile: 82.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `terminalbenchHard`.

82.5% percentile inside its fair comparison set

31.8%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 #10 · Source label: Gemini 3 Flash Preview (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 49.9%
Percentile: 97.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `scicode`.

97.8% percentile inside its fair comparison set

49.9%Raw benchmark value

Code 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 #24

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,437
Percentile: 68.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: overall. Source rank: #31. Votes: 13277. Organization: google. License: Proprietary.

68.5% percentile inside its fair comparison set

1,437Raw benchmark valueCI 1,430 - 1,445

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 #24

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,437
Percentile: 68.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: webdev. Source rank: #31. Votes: 13277. Organization: google. License: Proprietary.

68.5% percentile inside its fair comparison set

1,437Raw benchmark valueCI 1,430 - 1,445

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 #25

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,440
Percentile: 67.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: webdev-html. Source rank: #32. Votes: 9253. Organization: google. License: Proprietary.

67.1% percentile inside its fair comparison set

1,440Raw benchmark valueCI 1,430 - 1,450

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 #23

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,431
Percentile: 62.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: webdev-react. Source rank: #29. Votes: 3993. Organization: google. License: Proprietary.

62.7% percentile inside its fair comparison set

1,431Raw benchmark valueCI 1,420 - 1,441

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 #22 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 6.4%
Percentile: 4.5%
Last updated: recent
Eligibility: headline eligible

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

4.5% percentile inside its fair comparison set

6.4%Raw benchmark valueCI 4.2% - 8.5%

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 #16 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 85.6%
Percentile: 83.3%
Last updated: recent
Eligibility: headline eligible

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

83.3% percentile inside its fair comparison set

85.6%Raw benchmark valueCI 83.6% - 87.6%

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 #17 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 0%
Percentile: 70%
Last updated: recent
Eligibility: headline eligible

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

70% percentile inside its fair comparison set

0%Raw benchmark valueCI 0% - 0%

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 #19 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 75%
Percentile: 66.7%
Last updated: recent
Eligibility: headline eligible

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

66.7% percentile inside its fair comparison set

75%Raw benchmark valueCI 71.2% - 78.8%

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 #15 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 53.9%
Percentile: 48.1%
Last updated: recent
Eligibility: headline eligible

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

48.1% percentile inside its fair comparison set

53.9%Raw benchmark valueCI 51.4% - 56.5%

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 #29 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 20.2%
Percentile: 42.9%
Last updated: recent
Eligibility: headline eligible

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

42.9% percentile inside its fair comparison set

20.2%Raw benchmark valueCI 12.5% - 27.9%

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 #26

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,509
Percentile: 92.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: coding. Source rank: #33. Votes: 6387. Organization: google. License: Proprietary.

92.2% percentile inside its fair comparison set

1,509Raw benchmark valueCI 1,501 - 1,517

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 #36

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,463
Percentile: 89.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: coding. Source rank: #44. Votes: 6387. Organization: google. License: Proprietary.

89.1% percentile inside its fair comparison set

1,463Raw benchmark valueCI 1,455 - 1,471

IOI

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 #7 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 39.1%
Percentile: 86.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: ioi; provider: Google.

86.4% percentile inside its fair comparison set

39.1%Raw benchmark valueCI 18.4% - 59.8%

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 #10

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,447
Percentile: 47.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: image_to_webdev. Source rank: #13. Votes: 4435. Organization: google. License: Proprietary.

47.1% percentile inside its fair comparison set

1,447Raw benchmark valueCI 1,436 - 1,457

Reasoning / math / science8 benchmarks82%

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 #61 · Source label: Gemini 3 Flash Preview (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 14.1%
Percentile: 83.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `hle`.

83.8% percentile inside its fair comparison set

14.1%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 #57 · Source label: Gemini 3 Flash Preview (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 81.2%
Percentile: 85%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `gpqa`.

85% percentile inside its fair comparison set

81.2%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 #39 · Source label: Gemini 3 Flash Preview (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 1.4%
Percentile: 88.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `critpt`.

88.1% percentile inside its fair comparison set

1.4%Raw benchmark value

ProofBench

VALS-AI · Reasoning / math / science · Objective

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

Rank #23 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 15%
Percentile: 37.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: proof_bench; provider: Google.

37.1% percentile inside its fair comparison set

15%Raw benchmark valueCI 8% - 22%

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 #19 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 87.9%
Percentile: 79.8%
Last updated: recent
Eligibility: headline eligible

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

79.8% percentile inside its fair comparison set

87.9%Raw benchmark valueCI 84.7% - 91.1%

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 #9 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 88.6%
Percentile: 91%
Last updated: recent
Eligibility: headline eligible

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

91% percentile inside its fair comparison set

88.6%Raw benchmark valueCI 88% - 89.2%

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 #16

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,476
Percentile: 95.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: math. Source rank: #20. Votes: 2002. Organization: google. License: Proprietary.

95.2% percentile inside its fair comparison set

1,476Raw benchmark valueCI 1,463 - 1,489

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 #14

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,474
Percentile: 95.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: math. Source rank: #18. Votes: 2002. Organization: google. License: Proprietary.

95.9% percentile inside its fair comparison set

1,474Raw benchmark valueCI 1,460 - 1,487

Professional reasoning27 benchmarks84.5%

APEX-Agents-AA

AA · Professional reasoning · Objective

Long-horizon agentic task completion.

Rank #9 · Source label: Gemini 3 Flash Preview (Reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 27.7%
Percentile: 66.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `apexAgents`.

66.7% percentile inside its fair comparison set

27.7%Raw benchmark value

Vals Index

VALS-AI · Professional reasoning · Combined

Weighted model performance across economically relevant Vals tasks.

Rank #18 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 50
Percentile: 34.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: vals_index; provider: Google.

34.6% percentile inside its fair comparison set

50Raw benchmark valueCI 47 - 53

LegalBench

VALS-AI · Professional reasoning · Objective

Academic legal reasoning tasks.

Rank #4 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 86.9%
Percentile: 96.7%
Last updated: recent
Eligibility: headline eligible

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

96.7% percentile inside its fair comparison set

86.9%Raw benchmark valueCI 86.2% - 87.6%

Finance Agent v2

VALS-AI · Professional reasoning · Objective

Core financial analyst tasks for agentic models.

Rank #14 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 42.6%
Percentile: 48%
Last updated: recent
Eligibility: headline eligible

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

48% percentile inside its fair comparison set

42.6%Raw benchmark valueCI 41.7% - 43.4%

TaxEval v2

VALS-AI · Professional reasoning · Objective

Answer quality on tax questions and responses.

Rank #28 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 73.9%
Percentile: 70.3%
Last updated: recent
Eligibility: headline eligible

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

70.3% percentile inside its fair comparison set

73.9%Raw benchmark valueCI 72.2% - 75.6%

MedCode

VALS-AI · Professional reasoning · Objective

Medical billing support and coding tasks.

Rank #3 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 55.9%
Percentile: 96.1%
Last updated: recent
Eligibility: headline eligible

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

96.1% percentile inside its fair comparison set

55.9%Raw benchmark valueCI 51.8% - 60.1%

MedScribe

VALS-AI · Professional reasoning · Objective

Administrative documentation support for doctors.

Rank #45 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 69.9%
Percentile: 12%
Last updated: recent
Eligibility: headline eligible

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

12% percentile inside its fair comparison set

69.9%Raw benchmark valueCI 66.2% - 73.6%

Text Arena · Expert

AR · Professional reasoning · Human

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

Rank #19

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,497
Percentile: 93.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: expert. Source rank: #25. Votes: 1932. Organization: google. License: Proprietary.

93.5% percentile inside its fair comparison set

1,497Raw benchmark valueCI 1,483 - 1,510

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 #17

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,470
Percentile: 95%
Last updated: recent
Eligibility: headline eligible

95% percentile inside its fair comparison set

1,470Raw benchmark valueCI 1,462 - 1,479

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 #11

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,454
Percentile: 96.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: industry_entertainment_and_sports_and_media. Source rank: #13. Votes: 5585. Organization: google. License: Proprietary.

96.9% percentile inside its fair comparison set

1,454Raw benchmark valueCI 1,445 - 1,462

Text Arena · Industry Legal And Government

AR · Professional reasoning · Human

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

Rank #10

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,487
Percentile: 97%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: industry_legal_and_government. Source rank: #12. Votes: 2310. Organization: google. License: Proprietary.

97% percentile inside its fair comparison set

1,487Raw benchmark valueCI 1,475 - 1,500

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 #21

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,486
Percentile: 93.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: industry_life_and_physical_and_social_science. Source rank: #25. Votes: 4888. Organization: google. License: Proprietary.

93.8% percentile inside its fair comparison set

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

Text Arena · Industry Mathematical

AR · Professional reasoning · Human

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

Rank #24

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,472
Percentile: 92.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: industry_mathematical. Source rank: #28. Votes: 1414. Organization: google. License: Proprietary.

92.5% percentile inside its fair comparison set

1,472Raw benchmark valueCI 1,456 - 1,488

Text Arena · Industry Medicine And Healthcare

AR · Professional reasoning · Human

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

Rank #23

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,484
Percentile: 92.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: industry_medicine_and_healthcare. Source rank: #29. Votes: 2069. Organization: google. License: Proprietary.

92.5% percentile inside its fair comparison set

1,484Raw benchmark valueCI 1,471 - 1,498

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 #21

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,501
Percentile: 93.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: industry_software_and_it_services. Source rank: #27. Votes: 10515. Organization: google. License: Proprietary.

93.8% percentile inside its fair comparison set

1,501Raw benchmark valueCI 1,494 - 1,508

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 #14

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,461
Percentile: 96%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: industry_writing_and_literature_and_language. Source rank: #18. Votes: 6818. Organization: google. License: Proprietary.

96% percentile inside its fair comparison set

1,461Raw benchmark valueCI 1,453 - 1,469

Text Arena · Expert · No Style Control

AR · Professional reasoning · Human

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

Rank #28

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,463
Percentile: 90.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: expert. Source rank: #36. Votes: 1932. Organization: google. License: Proprietary.

90.2% percentile inside its fair comparison set

1,463Raw benchmark valueCI 1,450 - 1,477

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 #17

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,450
Percentile: 95%
Last updated: recent
Eligibility: headline eligible

95% percentile inside its fair comparison set

1,450Raw benchmark valueCI 1,442 - 1,458

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 #9

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,449
Percentile: 97.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: industry_entertainment_and_sports_and_media. Source rank: #11. Votes: 5585. Organization: google. License: Proprietary.

97.5% percentile inside its fair comparison set

1,449Raw benchmark valueCI 1,440 - 1,457

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 #13

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,477
Percentile: 96%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: industry_legal_and_government. Source rank: #15. Votes: 2310. Organization: google. License: Proprietary.

96% percentile inside its fair comparison set

1,477Raw benchmark valueCI 1,465 - 1,490

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 #18

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,472
Percentile: 94.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: industry_life_and_physical_and_social_science. Source rank: #22. Votes: 4888. Organization: google. License: Proprietary.

94.7% percentile inside its fair comparison set

1,472Raw benchmark valueCI 1,464 - 1,481

Text Arena · Industry Mathematical · No Style Control

AR · Professional reasoning · Human

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

Rank #22

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,465
Percentile: 93.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: industry_mathematical. Source rank: #26. Votes: 1414. Organization: google. License: Proprietary.

93.2% percentile inside its fair comparison set

1,465Raw benchmark valueCI 1,450 - 1,481

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 #22

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,465
Percentile: 92.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: industry_medicine_and_healthcare. Source rank: #25. Votes: 2069. Organization: google. License: Proprietary.

92.9% percentile inside its fair comparison set

1,465Raw benchmark valueCI 1,452 - 1,478

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 #24

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,471
Percentile: 92.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: industry_software_and_it_services. Source rank: #30. Votes: 10515. Organization: google. License: Proprietary.

92.9% percentile inside its fair comparison set

1,471Raw benchmark valueCI 1,465 - 1,478

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 #13

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,459
Percentile: 96.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: industry_writing_and_literature_and_language. Source rank: #16. Votes: 6818. Organization: google. License: Proprietary.

96.3% percentile inside its fair comparison set

1,459Raw benchmark valueCI 1,451 - 1,467

SAGE

VALS-AI · Professional reasoning · Objective

Student Assessment with Generative Evaluation.

Rank #6 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 51.8%
Percentile: 88.9%
Last updated: recent
Eligibility: headline eligible

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

88.9% percentile inside its fair comparison set

51.8%Raw benchmark valueCI 45.2% - 58.5%

Poker Agent

VALS-AI · Professional reasoning · Objective

Agent profit in poker-style strategic play.

Rank #6 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 1,100.2 score
Percentile: 68.8%
Last updated: archived
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: poker_agent; provider: unknown.

68.8% percentile inside its fair comparison set

1,100.2 scoreRaw benchmark valueCI 1,100.2 score - 1,100.2 score

Search / tool use3 benchmarks62.1%

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 #146 · Source label: Gemini 3 Flash Preview (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 43.3%
Percentile: 53.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `tau2`.

53.1% percentile inside its fair comparison set

43.3%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 #10 · Source label: gemini-3-flash-grounding

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,194
Percentile: 70%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash-grounding`. Category: overall. Source rank: #11. Votes: 93980. Organization: google. License: Proprietary.

70% percentile inside its fair comparison set

1,194Raw benchmark valueCI 1,189 - 1,199

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 #12 · Source label: gemini-3-flash-grounding

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,201
Percentile: 63.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash-grounding`. Category: overall. Source rank: #12. Votes: 93980. Organization: google. License: Proprietary.

63.3% percentile inside its fair comparison set

1,201Raw benchmark valueCI 1,196 - 1,206

Long context2 benchmarks78.4%

Long Context Reasoning

AA · Long context · Objective

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

Rank #98 · Source label: Gemini 3 Flash Preview (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 48%
Percentile: 69.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `lcr`.

69.2% percentile inside its fair comparison set

48%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 #12 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 66.4%
Percentile: 87.5%
Last updated: recent
Eligibility: headline eligible

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

87.5% percentile inside its fair comparison set

66.4%Raw benchmark valueCI 64.6% - 68.3%

Vision understanding22 benchmarks88.9%

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 #7 · Source label: Gemini 3 Flash Preview (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 78.6%
Percentile: 95.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `mmmuPro`.

95.6% percentile inside its fair comparison set

78.6%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 #14

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,272
Percentile: 88.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: overall. Source rank: #18. Votes: 28104. Organization: google. License: Proprietary.

88.1% percentile inside its fair comparison set

1,272Raw benchmark valueCI 1,266 - 1,278

Vision Arena · Captioning

AR · Vision understanding · Human

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

Rank #6

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,235
Percentile: 80.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: captioning. Source rank: #6. Votes: 163. Organization: google. License: Proprietary.

80.8% percentile inside its fair comparison set

1,235Raw benchmark valueCI 1,192 - 1,277

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 #6 · Source label: gemini-3-flash (thinking-minimal)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,283
Percentile: 90.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash (thinking-minimal)`. Category: creative_writing_vision. Source rank: #8. Votes: 1624. Organization: google. License: Proprietary.

90.9% percentile inside its fair comparison set

1,283Raw benchmark valueCI 1,268 - 1,298

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 #14

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,296
Percentile: 81.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: diagram. Source rank: #19. Votes: 7012. Organization: google. License: Proprietary.

81.4% percentile inside its fair comparison set

1,296Raw benchmark valueCI 1,288 - 1,305

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 #16

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,261
Percentile: 86.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: english. Source rank: #20. Votes: 11898. Organization: google. License: Proprietary.

86.2% percentile inside its fair comparison set

1,261Raw benchmark valueCI 1,253 - 1,269

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 #2

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,296
Percentile: 96.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: entity_recognition. Source rank: #2. Votes: 233. Organization: google. License: Proprietary.

96.9% percentile inside its fair comparison set

1,296Raw benchmark valueCI 1,260 - 1,332

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 #7

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,312
Percentile: 91.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: homework. Source rank: #10. Votes: 4223. Organization: google. License: Proprietary.

91.2% percentile inside its fair comparison set

1,312Raw benchmark valueCI 1,302 - 1,322

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 #9

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,268
Percentile: 83.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: humor. Source rank: #12. Votes: 958. Organization: google. License: Proprietary.

83.7% percentile inside its fair comparison set

1,268Raw benchmark valueCI 1,248 - 1,287

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 #13

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,286
Percentile: 82.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: ocr. Source rank: #17. Votes: 19274. Organization: google. License: Proprietary.

82.9% percentile inside its fair comparison set

1,286Raw benchmark valueCI 1,280 - 1,291

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 #11

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,285
Percentile: 90.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: overall. Source rank: #15. Votes: 28104. Organization: google. License: Proprietary.

90.8% percentile inside its fair comparison set

1,285Raw benchmark valueCI 1,279 - 1,290

Vision Arena · Captioning · 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 #5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,282
Percentile: 84.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: captioning. Source rank: #5. Votes: 163. Organization: google. License: Proprietary.

84.6% percentile inside its fair comparison set

1,282Raw benchmark valueCI 1,240 - 1,323

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 #6

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,297
Percentile: 90.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: creative_writing_vision. Source rank: #8. Votes: 1817. Organization: google. License: Proprietary.

90.9% percentile inside its fair comparison set

1,297Raw benchmark valueCI 1,283 - 1,311

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

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,294
Percentile: 84.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: diagram. Source rank: #17. Votes: 7012. Organization: google. License: Proprietary.

84.3% percentile inside its fair comparison set

1,294Raw benchmark valueCI 1,286 - 1,302

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 #15

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,274
Percentile: 87.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: english. Source rank: #19. Votes: 11898. Organization: google. License: Proprietary.

87.2% percentile inside its fair comparison set

1,274Raw benchmark valueCI 1,266 - 1,282

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 #2

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,338
Percentile: 96.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: entity_recognition. Source rank: #2. Votes: 233. Organization: google. License: Proprietary.

96.9% percentile inside its fair comparison set

1,338Raw benchmark valueCI 1,302 - 1,375

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 #9

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,317
Percentile: 88.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: homework. Source rank: #12. Votes: 4223. Organization: google. License: Proprietary.

88.2% percentile inside its fair comparison set

1,317Raw benchmark valueCI 1,307 - 1,327

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 #6

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,289
Percentile: 89.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: humor. Source rank: #8. Votes: 958. Organization: google. License: Proprietary.

89.8% percentile inside its fair comparison set

1,289Raw benchmark valueCI 1,270 - 1,309

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 #12

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,291
Percentile: 84.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: ocr. Source rank: #16. Votes: 19274. Organization: google. License: Proprietary.

84.3% percentile inside its fair comparison set

1,291Raw benchmark valueCI 1,286 - 1,297

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 #5 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 87.6%
Percentile: 93.1%
Last updated: recent
Eligibility: headline eligible

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

93.1% percentile inside its fair comparison set

87.6%Raw benchmark valueCI 86.1% - 89.2%

Vision Arena · Creative Writing

AR · Vision understanding · Human

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

Rank #3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,289
Percentile: 93.8%
Last updated: archived
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: creative_writing. Source rank: #3. Votes: 217. Organization: google. License: Proprietary.

93.8% percentile inside its fair comparison set

1,289Raw benchmark valueCI 1,249 - 1,329

Vision Arena · Creative Writing · 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 #3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,306
Percentile: 93.8%
Last updated: archived
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: creative_writing. Source rank: #3. Votes: 217. Organization: google. License: Proprietary.

93.8% percentile inside its fair comparison set

1,306Raw benchmark valueCI 1,267 - 1,346

Document understanding4 benchmarks43.8%

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 #17

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,442
Percentile: 33.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: overall. Source rank: #20. Votes: 7188. Organization: google. License: Proprietary.

33.3% percentile inside its fair comparison set

1,442Raw benchmark valueCI 1,433 - 1,451

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 #21

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,413
Percentile: 16.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: overall. Source rank: #24. Votes: 7188. Organization: google. License: Proprietary.

16.7% percentile inside its fair comparison set

1,413Raw benchmark valueCI 1,404 - 1,422

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 #13 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 52
Percentile: 36.8%
Last updated: recent
Eligibility: headline eligible

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

36.8% percentile inside its fair comparison set

52Raw benchmark valueCI 49 - 55

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 #8 · Source label: google/gemini-3-flash-preview

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 68.7%
Percentile: 88.3%
Last updated: recent
Eligibility: headline eligible

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

88.3% percentile inside its fair comparison set

68.7%Raw benchmark valueCI 66.9% - 70.5%

Multilingual16 benchmarks95.4%

Text Arena · Chinese

AR · Multilingual · Human

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

Rank #11

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,521
Percentile: 96.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: chinese. Source rank: #15. Votes: 1513. Organization: google. License: Proprietary.

96.6% percentile inside its fair comparison set

1,521Raw benchmark valueCI 1,506 - 1,537

Text Arena · French

AR · Multilingual · Human

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

Rank #8

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,506
Percentile: 96.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: french. Source rank: #9. Votes: 754. Organization: google. License: Proprietary.

96.8% percentile inside its fair comparison set

1,506Raw benchmark valueCI 1,483 - 1,529

Text Arena · German

AR · Multilingual · Human

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

Rank #6

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,494
Percentile: 97.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: german. Source rank: #7. Votes: 531. Organization: google. License: Proprietary.

97.9% percentile inside its fair comparison set

1,494Raw benchmark valueCI 1,468 - 1,520

Text Arena · Japanese

AR · Multilingual · Human

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

Rank #6

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,492
Percentile: 97.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: japanese. Source rank: #6. Votes: 272. Organization: google. License: Proprietary.

97.5% percentile inside its fair comparison set

1,492Raw benchmark valueCI 1,454 - 1,530

Text Arena · Korean

AR · Multilingual · Human

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

Rank #12

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,447
Percentile: 94.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: korean. Source rank: #13. Votes: 537. Organization: google. License: Proprietary.

94.7% percentile inside its fair comparison set

1,447Raw benchmark valueCI 1,421 - 1,474

Text Arena · Russian

AR · Multilingual · Human

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

Rank #10

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,486
Percentile: 96.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: russian. Source rank: #13. Votes: 3264. Organization: google. License: Proprietary.

96.9% percentile inside its fair comparison set

1,486Raw benchmark valueCI 1,475 - 1,496

Text Arena · Spanish

AR · Multilingual · Human

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

Rank #15

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,468
Percentile: 93.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: spanish. Source rank: #17. Votes: 832. Organization: google. License: Proprietary.

93.5% percentile inside its fair comparison set

1,468Raw benchmark valueCI 1,447 - 1,489

Text Arena · Chinese · No Style Control

AR · Multilingual · Human

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

Rank #12

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,514
Percentile: 96.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: chinese. Source rank: #16. Votes: 1513. Organization: google. License: Proprietary.

96.3% percentile inside its fair comparison set

1,514Raw benchmark valueCI 1,498 - 1,529

Text Arena · French · No Style Control

AR · Multilingual · Human

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

Rank #10

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,493
Percentile: 95.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: french. Source rank: #12. Votes: 754. Organization: google. License: Proprietary.

95.8% percentile inside its fair comparison set

1,493Raw benchmark valueCI 1,470 - 1,516

Text Arena · German · No Style Control

AR · Multilingual · Human

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

Rank #5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,493
Percentile: 98.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: german. Source rank: #5. Votes: 531. Organization: google. License: Proprietary.

98.3% percentile inside its fair comparison set

1,493Raw benchmark valueCI 1,467 - 1,519

Text Arena · Japanese · No Style Control

AR · Multilingual · Human

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

Rank #5

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,495
Percentile: 98%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: japanese. Source rank: #5. Votes: 272. Organization: google. License: Proprietary.

98% percentile inside its fair comparison set

1,495Raw benchmark valueCI 1,456 - 1,533

Text Arena · Korean · No Style Control

AR · Multilingual · Human

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

Rank #9

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,446
Percentile: 96.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: korean. Source rank: #9. Votes: 537. Organization: google. License: Proprietary.

96.2% percentile inside its fair comparison set

1,446Raw benchmark valueCI 1,419 - 1,473

Text Arena · Russian · No Style Control

AR · Multilingual · Human

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

Rank #10

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,479
Percentile: 96.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: russian. Source rank: #13. Votes: 3264. Organization: google. License: Proprietary.

96.9% percentile inside its fair comparison set

1,479Raw benchmark valueCI 1,468 - 1,489

Text Arena · Spanish · No Style Control

AR · Multilingual · Human

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

Rank #17

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,460
Percentile: 92.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: spanish. Source rank: #19. Votes: 832. Organization: google. License: Proprietary.

92.5% percentile inside its fair comparison set

1,460Raw benchmark valueCI 1,439 - 1,481

Vision Arena · Chinese

AR · Multilingual · Human

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

Rank #8

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,323
Percentile: 90.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: chinese. Source rank: #10. Votes: 1512. Organization: google. License: Proprietary.

90.9% percentile inside its fair comparison set

1,323Raw benchmark valueCI 1,303 - 1,343

Vision Arena · Chinese · No Style Control

AR · Multilingual · Human

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

Rank #11

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,337
Percentile: 87%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `gemini-3-flash`. Category: chinese. Source rank: #15. Votes: 1512. Organization: google. License: Proprietary.

87% percentile inside its fair comparison set

1,337Raw benchmark valueCI 1,317 - 1,356