supermarket

Vision-language

Claude Opus 4.7

By Anthropic, in the Claude Opus family.

Released
16 Apr 2026
Status
Legacy
Context
1M
Max output
128K
Input per 1M tokens
$5.00
Output per 1M tokens
$25.00
Open weights
No
License
proprietary
Parameters
—
Knowledge cutoff
Jan 2026
Reasoning
Yes
Modalities
text, image → text

Category scores

Out of 100, from independent and third-party results only, by method category-2026-09-29.1.

Category scores
CategoryScoreRankCoverage
Knowledge & reasoning40.2364 of 4
Coding & software engineering37.973 of 5
Computer & browser use18.241 of 1
Vision & multimodal understanding4.8Not ranked1 of 3
Factuality & hallucination51.7201 of 1
Human preference52.0255 of 5

Benchmark results

Every result we hold for this model, with who ran it and where it was published.

Knowledge & reasoning

Knowledge & reasoning results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
Chess PuzzlesPuzzles solved30.0%Independentxhigh20 Apr 2026Epoch AI
Chess PuzzlesPuzzles solved20.0%Independentlow14 Jul 2026Epoch AI
Chess PuzzlesPuzzles solved7.0%Independentmax6 Aug 2026Epoch AI
DTBench (Decision Theory Benchmark)Accuracy (capability questions)94.7%Third partymax—https://conceptualreasoning.ai/dtbench via Epoch AI
Earthborne Rangers (EBR-bench)Objectives reached19.0%Independentmax30 Jun 2026Epoch AI
LMCA (Language Model Conceptual Argumentation)Correlation with expert ratings (Pearson r)0.522Third partymax—https://conceptualreasoning.ai/lmca via Epoch AI
Mystery Game PuzzlesPuzzles solved28.0%Independentmax25 Jul 2026Epoch AI
Mystery Game PuzzlesPuzzles solved13.0%Independent—6 Aug 2026Epoch AI
SimpleBenchScore (AVG@5)61.7%Third partyunknown—https://lmcouncil.ai/benchmarks via Epoch AI

Scientific reasoning

Scientific reasoning results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
GPQA DiamondAccuracy90.2%Independentxhigh17 Apr 2026Epoch AI
GPQA DiamondAccuracy86.4%Independentmax6 Aug 2026Epoch AI

Mathematics

Mathematics results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
FrontierMath Tier 4Accuracy31.7%Independentmax10 Jun 2026Epoch AI
FrontierMath Tier 4 (v1)Accuracy22.9%Independentxhigh17 Apr 2026Epoch AI
FrontierMath Tiers 1-3Accuracy70.2%Independentmax10 Jun 2026Epoch AI
FrontierMath Tiers 1-3 (v1)Accuracy43.8%Independentxhigh17 Apr 2026Epoch AI
OTIS Mock AIME 2024-2025Accuracy97.8%Independentxhigh17 Apr 2026Epoch AI
OTIS Mock AIME 2024-2025Accuracy86.7%Independentmax6 Aug 2026Epoch AI

Coding & software engineering

Coding & software engineering results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
ExploitBenchMean capability26.5%Source not stated——Epoch AI via Epoch AI
ExploitBenchMean capability23.9%Source not stated——Epoch AI via Epoch AI
FrontierCodeMain score (Mean@5)38.5%Third partyunknown—https://cognition.com/frontiercode via Epoch AI
GSOOPT@144.1%Third partyhigh—https://gso-bench.github.io/index.html via Epoch AI
GSOOPT@144.1%Third party——GSO Leaderboard via Epoch AI
MirrorCodeScore31.1%Independenthigh12 Aug 2026Epoch AI
PostTrainBenchAverage score28.6%Source not statedxhigh—Epoch AI via Epoch AI
SWE-bench Verified% resolved83.5%Independentmax20 Apr 2026Epoch AI
Terminal-Bench 2.0Accuracy (mean over trials)80.2%Third partyunknown—https://www.tbench.ai/leaderboard/terminal-bench/2.0 via Epoch AI
WeirdML (v2)Average accuracy76.4%Third partyhigh—https://htihle.github.io/weirdml.html via Epoch AI
WeirdML (v2)Average accuracy76.4%Third partyunknown—WeirdML Leaderboard via Epoch AI
WeirdML (v2)Average accuracy76.4%Third party——https://htihle.github.io/weirdml.html via Epoch AI
WeirdML (v2)Average accuracy75.5%Third partymax—https://htihle.github.io/weirdml.html via Epoch AI

Computer & browser use

Computer & browser use results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
OSWorld 2.0Binary completion (500 steps)18.2%Third partymax—https://osworld-v2.xlang.ai/ via Epoch AI
OSWorld 2.0Binary completion (500 steps)13.9%Third partymax—https://osworld-v2.xlang.ai/ via Epoch AI

Professional knowledge work

Professional knowledge work results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
APEX-AgentsPass@1 (all rubric criteria met)49.2%Source not statedmax—Epoch AI via Epoch AI

Vision & multimodal understanding

Vision & multimodal understanding results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
Furniture AssemblyScore33.3%Independentmax10 Sep 2026Epoch AI

Factuality & hallucination

Factuality & hallucination results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
SimpleQA VerifiedAccuracy (% correct)51.7%Independentxhigh27 Aug 2026Epoch AI

Human preference

Human preference results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
Code Arena: WebDev (arena.ai)Arena score (Elo scale)1557Independenthigh25 Sep 2026Arena
Code Arena: WebDev (arena.ai)Arena score (Elo scale)1556Independent—25 Sep 2026Arena
Code Arena: WebDev (arena.ai)Votes17538Independenthigh25 Sep 2026Arena
Code Arena: WebDev (arena.ai)Votes16945Independent—25 Sep 2026Arena
Document Arena (arena.ai)Arena score (Elo scale)1494Independent—13 Sep 2026Arena
Document Arena (arena.ai)Arena score (Elo scale)1493Independenthigh13 Sep 2026Arena
Document Arena (arena.ai)Votes22192Independent—13 Sep 2026Arena
Document Arena (arena.ai)Votes21957Independenthigh13 Sep 2026Arena
Search Arena (arena.ai)Arena score (Elo scale)1212Independent—24 Aug 2026Arena
Search Arena (arena.ai)Votes91394Independent—24 Aug 2026Arena
Text Arena (arena.ai)Arena score (Elo scale)1502Independenthigh25 Sep 2026Arena
Text Arena (arena.ai)Arena score (Elo scale)1495Independent—25 Sep 2026Arena
Text Arena (arena.ai)Votes65051Independent—25 Sep 2026Arena
Text Arena (arena.ai)Votes64007Independenthigh25 Sep 2026Arena
Vision Arena (arena.ai)Arena score (Elo scale)1299Independent—28 Sep 2026Arena
Vision Arena (arena.ai)Arena score (Elo scale)1298Independenthigh28 Sep 2026Arena
Vision Arena (arena.ai)Votes22948Independent—28 Sep 2026Arena
Vision Arena (arena.ai)Votes22559Independenthigh28 Sep 2026Arena

Other Claude Opus versions

Sources