UAB

Claude Haiku 4.5

Live · updated continuously

MenuClaude Haiku 4.5

Find

Find a model Comparison table Score overview

Browse

Models Benchmarks Sources Compare

Method

Data status Data & sources Scoring method Methodology Changelog

Model profile · Anthropic

Claude Haiku 4.5

Closed weightsmid · registry tag 2026 fast

Thin verified coverage

Reads as thin verified coverage across the resolved source data.

Visible coverage: 34.2%
Verified coverage: 32.2%
Spread: 95.8%
Last verified: Aug 19, 2026

39%bench fit

textvisiondocumentcode7 aliases37 official source links

Open compare

UAB

Live · updated continuously

MenuLoading

Find

Find a model Comparison table Score overview

Browse

Models Benchmarks Sources Compare

Method

Data status Data & sources Scoring method Methodology Changelog

Model detail

Loading model evidence.

Preparing the model profile, benchmark coverage, and registry evidence.

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 / text29 benchmarks67.4%

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 #132 · Source label: Claude 4.5 Haiku (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 24
Percentile: 72%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `intelligenceIndex`.

72% percentile inside its fair comparison set

24Raw 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 #262 · Source label: Claude 4.5 Haiku (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 14.4%
Percentile: 29%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `omniscienceAccuracy`.

29% percentile inside its fair comparison set

14.4%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 #10 · Source label: Claude 4.5 Haiku (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 74.3%
Percentile: 97.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `omniscienceNonHallucination`.

97.5% percentile inside its fair comparison set

74.3%Raw benchmark value

IFBench

AA · Chat / text · Objective

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

Rank #167 · Source label: Claude 4.5 Haiku (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 42%
Percentile: 52.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `ifbench`.

52.2% percentile inside its fair comparison set

42%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 #219 · Source label: Claude 4.5 Haiku (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: $2 /1M tokens
Percentile: 28.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `price1mBlended0To3To1`.

28.2% percentile inside its fair comparison set

$2 /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 #210 · Source label: Claude 4.5 Haiku (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: $1 /1M input tokens
Percentile: 32.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `price1mInputTokens`.

32.2% percentile inside its fair comparison set

$1 /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 #226 · Source label: Claude 4.5 Haiku (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: $5 /1M output tokens
Percentile: 25.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `price1mOutputTokens`.

25.9% percentile inside its fair comparison set

$5 /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 #111 · Source label: Claude 4.5 Haiku (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 90.1 tokens/s
Percentile: 50.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `medianOutputTokensPerSecond`.

50.9% percentile inside its fair comparison set

90.1 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 #39 · Source label: Claude 4.5 Haiku (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 1.04s
Percentile: 83%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `medianTimeToFirstTokenSeconds`.

83% percentile inside its fair comparison set

1.04sRaw 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 #27 · Source label: Claude 4.5 Haiku (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 1.04s
Percentile: 88.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `medianTimeToFirstAnswerTokenSeconds`.

88.4% percentile inside its fair comparison set

1.04sRaw benchmark value

Openness Index

AA · Chat / text · Combined

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

Rank #240 · Source label: Claude 4.5 Haiku (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 11
Percentile: 11.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `opennessBreakdown.opennessIndex`.

11.3% percentile inside its fair comparison set

11Raw 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 #86 · Source label: claude-haiku-4-5-20251001

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,411
Percentile: 73.8%
Last updated: aging
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `claude-haiku-4-5-20251001`. Category: overall. Source rank: #106. Votes: 91153. Organization: anthropic. License: Proprietary.

73.8% percentile inside its fair comparison set

1,411Raw benchmark valueCI 1,408 - 1,414

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 #78 · Source label: claude-haiku-4-5-20251001

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,386
Percentile: 76.2%
Last updated: aging
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `claude-haiku-4-5-20251001`. Category: creative_writing. Source rank: #96. Votes: 13641. Organization: anthropic. License: Proprietary.

76.2% percentile inside its fair comparison set

1,386Raw benchmark valueCI 1,381 - 1,392

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 #71 · Source label: claude-haiku-4-5-20251001

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,432
Percentile: 78.5%
Last updated: aging
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `claude-haiku-4-5-20251001`. Category: english. Source rank: #89. Votes: 43281. Organization: anthropic. License: Proprietary.

78.5% percentile inside its fair comparison set

1,432Raw benchmark valueCI 1,428 - 1,435

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 #86 · Source label: claude-haiku-4-5-20251001

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,398
Percentile: 73.8%
Last updated: aging
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `claude-haiku-4-5-20251001`. Category: exclude_ties. Source rank: #106. Votes: 67483. Organization: anthropic. License: Proprietary.

73.8% percentile inside its fair comparison set

1,398Raw benchmark valueCI 1,394 - 1,402

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 #73 · Source label: claude-haiku-4-5-20251001

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,439
Percentile: 77.8%
Last updated: aging
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `claude-haiku-4-5-20251001`. Category: hard_prompts. Source rank: #91. Votes: 53244. Organization: anthropic. License: Proprietary.

77.8% percentile inside its fair comparison set

1,439Raw benchmark valueCI 1,435 - 1,443

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 #63 · Source label: claude-haiku-4-5-20251001

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,455
Percentile: 80.9%
Last updated: aging
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `claude-haiku-4-5-20251001`. Category: hard_prompts_english. Source rank: #79. Votes: 26560. Organization: anthropic. License: Proprietary.

80.9% percentile inside its fair comparison set

1,455Raw benchmark valueCI 1,450 - 1,459

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 #69 · Source label: claude-haiku-4-5-20251001

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,414
Percentile: 79.1%
Last updated: aging
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `claude-haiku-4-5-20251001`. Category: instruction_following. Source rank: #86. Votes: 26995. Organization: anthropic. License: Proprietary.

79.1% percentile inside its fair comparison set

1,414Raw benchmark valueCI 1,410 - 1,419

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 #61 · Source label: claude-haiku-4-5-20251001

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,437
Percentile: 80.3%
Last updated: aging
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `claude-haiku-4-5-20251001`. Category: longer_query. Source rank: #76. Votes: 29096. Organization: anthropic. License: Proprietary.

80.3% percentile inside its fair comparison set

1,437Raw benchmark valueCI 1,432 - 1,441

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 #68 · Source label: claude-haiku-4-5-20251001

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,426
Percentile: 79.3%
Last updated: aging
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `claude-haiku-4-5-20251001`. Category: multi_turn. Source rank: #86. Votes: 16101. Organization: anthropic. License: Proprietary.

79.3% percentile inside its fair comparison set

1,426Raw benchmark valueCI 1,420 - 1,431

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 #101 · Source label: claude-haiku-4-5-20251001

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,392
Percentile: 69.2%
Last updated: aging
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `claude-haiku-4-5-20251001`. Category: overall. Source rank: #122. Votes: 91153. Organization: anthropic. License: Proprietary.

69.2% percentile inside its fair comparison set

1,392Raw benchmark valueCI 1,389 - 1,395

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 #82 · Source label: claude-haiku-4-5-20251001

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,369
Percentile: 74.9%
Last updated: aging
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `claude-haiku-4-5-20251001`. Category: creative_writing. Source rank: #102. Votes: 13641. Organization: anthropic. License: Proprietary.

74.9% percentile inside its fair comparison set

1,369Raw benchmark valueCI 1,363 - 1,375

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 #92 · Source label: claude-haiku-4-5-20251001

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,413
Percentile: 72%
Last updated: aging
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `claude-haiku-4-5-20251001`. Category: english. Source rank: #108. Votes: 43281. Organization: anthropic. License: Proprietary.

72% percentile inside its fair comparison set

1,413Raw benchmark valueCI 1,409 - 1,417

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 #101 · Source label: claude-haiku-4-5-20251001

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,370
Percentile: 69.2%
Last updated: aging
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `claude-haiku-4-5-20251001`. Category: exclude_ties. Source rank: #122. Votes: 67483. Organization: anthropic. License: Proprietary.

69.2% percentile inside its fair comparison set

1,370Raw benchmark valueCI 1,367 - 1,374

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 #80 · Source label: claude-haiku-4-5-20251001

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,417
Percentile: 75.7%
Last updated: aging
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `claude-haiku-4-5-20251001`. Category: hard_prompts. Source rank: #97. Votes: 53244. Organization: anthropic. License: Proprietary.

75.7% percentile inside its fair comparison set

1,417Raw benchmark valueCI 1,413 - 1,420

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 #65 · Source label: claude-haiku-4-5-20251001

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,434
Percentile: 80.2%
Last updated: aging
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `claude-haiku-4-5-20251001`. Category: hard_prompts_english. Source rank: #80. Votes: 26560. Organization: anthropic. License: Proprietary.

80.2% percentile inside its fair comparison set

1,434Raw benchmark valueCI 1,429 - 1,439

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 #50 · Source label: claude-haiku-4-5-20251001

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,412
Percentile: 84.9%
Last updated: aging
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `claude-haiku-4-5-20251001`. Category: instruction_following. Source rank: #63. Votes: 26995. Organization: anthropic. License: Proprietary.

84.9% percentile inside its fair comparison set

1,412Raw benchmark valueCI 1,407 - 1,416

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 #49 · Source label: claude-haiku-4-5-20251001

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,427
Percentile: 84.2%
Last updated: aging
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `claude-haiku-4-5-20251001`. Category: longer_query. Source rank: #61. Votes: 29096. Organization: anthropic. License: Proprietary.

84.2% percentile inside its fair comparison set

1,427Raw benchmark valueCI 1,423 - 1,432

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 #81 · Source label: claude-haiku-4-5-20251001

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,408
Percentile: 75.2%
Last updated: aging
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `claude-haiku-4-5-20251001`. Category: multi_turn. Source rank: #98. Votes: 16101. Organization: anthropic. License: Proprietary.

75.2% percentile inside its fair comparison set

1,408Raw benchmark valueCI 1,403 - 1,414

Coding17 benchmarks37.5%

Terminal-Bench Hard

AA · Coding · Objective

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

Rank #77 · Source label: Claude 4.5 Haiku (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 27.3%
Percentile: 77.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `terminalbenchHard`.

77.2% percentile inside its fair comparison set

27.3%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 #184 · Source label: Claude 4.5 Haiku (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 34.4%
Percentile: 58.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `scicode`.

Reasoning / math / science7 benchmarks46.9%

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 #340 · Source label: Claude 4.5 Haiku (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 4.2%
Percentile: 23.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `hle`.

23.3% percentile inside its fair comparison set

4.2%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 #229 · Source label: Claude 4.5 Haiku (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 64.6%
Percentile: 49.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `gpqa`.

Professional reasoning29 benchmarks57.5%

GDPval-AA

AA · Professional reasoning · Rubric

Agentic performance on economically valuable work tasks.

Rank #137 · Source label: Claude 3.5 Haiku

backfilledproxy backfilledBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 457
Percentile: 20.6%
Last updated: recent
Eligibility: Fallback benchmark identity is visible for context but excluded from default ranking.

Parsed from Artificial Analysis public leaderboard field `gdpvalBreakdown.elo`. Backfilled from Claude Haiku 3.5 via approved benchmark identity mapping map-claude-haiku-4-5-to-3-5.

20.6% percentile inside its fair comparison set

457Raw benchmark value

Public Benefits Bench

VALS-AI · Professional reasoning · Objective

Answering SNAP benefits questions across the public-benefits lifecycle.

Rank #9 · Source label: anthropic/claude-haiku-4-5-20251001

verified runtimevariant direct

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 54.3%
Percentile: 27.3%
Last updated

Search / tool use1 benchmark47.3%

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 #178 · Source label: Claude 4.5 Haiku (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 32.5%
Percentile: 47.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `tau2`.

47.3% percentile inside its fair comparison set

32.5%Raw benchmark value

Long context2 benchmarks54.2%

Long Context Reasoning

AA · Long context · Objective

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

Rank #148 · Source label: Claude 4.5 Haiku (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 48.3%
Percentile: 61.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `lcr`.

61.7% percentile inside its fair comparison set

48.3%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 #48 · Source label: anthropic/claude-haiku-4-5-20251001-thinking

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 60.6%
Percentile: 46.6%
Last updated: aging
Eligibility: headline eligible

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

Vision understanding12 benchmarks14.7%

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 #115 · Source label: Claude 4.5 Haiku (Non-reasoning)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 55.1%
Percentile: 30.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `mmmuPro`.

30.5% percentile inside its fair comparison set

55.1%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 #75 · Source label: claude-3-5-haiku-20241022

backfilledproxy backfilledBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,126
Percentile: 33%
Last updated: aging
Eligibility: Fallback benchmark identity is visible for context but excluded from default ranking.

Document understanding4 benchmarks23.5%

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 #23 · Source label: claude-haiku-4-5-20251001

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,417
Percentile: 8.3%
Last updated: aging
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `claude-haiku-4-5-20251001`. Category: overall. Source rank: #27. Votes: 26371. Organization: anthropic. License: Proprietary.

8.3% percentile inside its fair comparison set

1,417Raw benchmark valueCI 1,410 - 1,423

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 #19 · Source label: claude-haiku-4-5-20251001

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,418
Percentile: 25%
Last updated: aging
Eligibility

Multilingual14 benchmarks65.6%

Text Arena · Chinese

AR · Multilingual · Human

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

Rank #88 · Source label: claude-haiku-4-5-20251001

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,436
Percentile: 70.5%
Last updated: aging
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `claude-haiku-4-5-20251001`. Category: chinese. Source rank: #110. Votes: 4491. Organization: anthropic. License: Proprietary.

70.5% percentile inside its fair comparison set

1,436Raw benchmark valueCI 1,426 - 1,445

Text Arena · French

AR · Multilingual · Human

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

Rank #82 · Source label: claude-haiku-4-5-20251001

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,425
Percentile: 62.5%
Last updated: aging
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `claude-haiku-4-5-20251001`. Category: french. Source rank: #100. Votes: 2195. Organization: anthropic. License: Proprietary.

Claude Haiku 4.5

Loading model evidence.

Claude Haiku 4.5

Source-linked scores by benchmark

Source links and registry checks

Current snapshot.