Updated daily at 05:00 UTC

Vision-language

Claude Sonnet 5

By Anthropic, in the Claude Sonnet family.

Released: 30 Jun 2026
Status: Legacy
Context: 1M
Max output: 128K
Input per 1M tokens: $2.00
Output per 1M tokens: $10.00
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
Category	Score	Rank	Coverage
Knowledge & reasoning	47.7	21	3 of 4
Coding & software engineering	44.6	5	3 of 5
Factuality & hallucination	33.7	47	1 of 1
Human preference	55.0	16	4 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
Benchmark	Metric	Value	Provenance	Setting	Evaluated	Source
Chess Puzzles	Puzzles solved	35.0%	Independent	xhigh	30 Jun 2026	Epoch AI
Chess Puzzles	Puzzles solved	16.0%	Independent	max	30 Jun 2026	Epoch AI
DTBench (Decision Theory Benchmark)	Accuracy (capability questions)	92.5%	Third party	max	—	https://conceptualreasoning.ai/dtbench via Epoch AI
DTBench (Decision Theory Benchmark)	Accuracy (capability questions)	89.5%	Third party	xhigh	—	https://conceptualreasoning.ai/dtbench via Epoch AI
DTBench (Decision Theory Benchmark)	Accuracy (capability questions)	84.5%	Third party	high	—	https://conceptualreasoning.ai/dtbench via Epoch AI
DTBench (Decision Theory Benchmark)	Accuracy (capability questions)	79.2%	Third party	medium	—	https://conceptualreasoning.ai/dtbench via Epoch AI
DTBench (Decision Theory Benchmark)	Accuracy (capability questions)	73.6%	Third party	low	—	https://conceptualreasoning.ai/dtbench via Epoch AI
LMCA (Language Model Conceptual Argumentation)	Correlation with expert ratings (Pearson r)	0.5002	Third party	high	—	https://conceptualreasoning.ai/lmca via Epoch AI
LMCA (Language Model Conceptual Argumentation)	Correlation with expert ratings (Pearson r)	0.4971	Third party	medium	—	https://conceptualreasoning.ai/lmca via Epoch AI
LMCA (Language Model Conceptual Argumentation)	Correlation with expert ratings (Pearson r)	0.4965	Third party	low	—	https://conceptualreasoning.ai/lmca via Epoch AI
LMCA (Language Model Conceptual Argumentation)	Correlation with expert ratings (Pearson r)	0.493	Third party	max	—	https://conceptualreasoning.ai/lmca via Epoch AI
LMCA (Language Model Conceptual Argumentation)	Correlation with expert ratings (Pearson r)	0.4881	Third party	xhigh	—	https://conceptualreasoning.ai/lmca via Epoch AI
Mystery Game Puzzles	Puzzles solved	35.0%	Independent	max	28 Jul 2026	Epoch AI
Mystery Game Puzzles	Puzzles solved	16.0%	Independent	—	6 Aug 2026	Epoch AI
SimpleBench	Score (AVG@5)	60.6%	Third party	unknown	—	SimpleBench Leaderboard via Epoch AI

Scientific reasoning

Scientific reasoning results
Benchmark	Metric	Value	Provenance	Setting	Evaluated	Source
GPQA Diamond	Accuracy	90.5%	Independent	xhigh	1 Jul 2026	Epoch AI
GPQA Diamond	Accuracy	80.3%	Independent	max	6 Aug 2026	Epoch AI
Surface Evolver Bench	Mean score (partial credit)	60.0%	Third party	medium	—	https://yhenon.github.io/surface-evolver-llm-eval/ via Epoch AI

Mathematics

Mathematics results
Benchmark	Metric	Value	Provenance	Setting	Evaluated	Source
FrontierMath Tier 4	Accuracy	29.3%	Independent	max	30 Jun 2026	Epoch AI
FrontierMath Tiers 1-3	Accuracy	65.6%	Independent	max	30 Jun 2026	Epoch AI
OTIS Mock AIME 2024-2025	Accuracy	94.7%	Independent	xhigh	1 Jul 2026	Epoch AI
OTIS Mock AIME 2024-2025	Accuracy	80.0%	Independent	max	6 Aug 2026	Epoch AI

Coding & software engineering

Coding & software engineering results
Benchmark	Metric	Value	Provenance	Setting	Evaluated	Source
DeepSWE	Pass@1	53.8%	Third party	max	—	https://deepswe.datacurve.ai/ via Epoch AI
DeepSWE	Pass@1	49.7%	Third party	xhigh	—	https://deepswe.datacurve.ai/ via Epoch AI
DeepSWE	Pass@1	48.2%	Third party	high	—	https://deepswe.datacurve.ai/ via Epoch AI
DeepSWE	Pass@1	39.8%	Third party	medium	—	https://deepswe.datacurve.ai/ via Epoch AI
DeepSWE	Pass@1	30.5%	Third party	low	—	https://deepswe.datacurve.ai/ via Epoch AI
FrontierCode	Main score (Mean@5)	42.7%	Third party	unknown	—	https://cognition.com/frontiercode via Epoch AI
GSO	OPT@1	37.3%	Third party	unknown	—	https://gso-bench.github.io/index.html via Epoch AI
WeirdML (v2)	Average accuracy	68.8%	Third party	high	—	https://htihle.github.io/weirdml.html via Epoch AI

Professional knowledge work

Professional knowledge work results
Benchmark	Metric	Value	Provenance	Setting	Evaluated	Source
APEX-Agents	Pass@1 (all rubric criteria met)	54.5%	Source not stated	unknown	—	Epoch AI via Epoch AI

Factuality & hallucination

Factuality & hallucination results
Benchmark	Metric	Value	Provenance	Setting	Evaluated	Source
SimpleQA Verified	Accuracy (% correct)	33.7%	Independent	max	27 Aug 2026	Epoch AI
SimpleQA Verified	Accuracy (% correct)	32.9%	Independent	xhigh	27 Aug 2026	Epoch AI

Human preference

Human preference results
Benchmark	Metric	Value	Provenance	Setting	Evaluated	Source
Code Arena: WebDev (arena.ai)	Arena score (Elo scale)	1539	Independent	high	25 Sep 2026	Arena
Code Arena: WebDev (arena.ai)	Votes	12862	Independent	high	25 Sep 2026	Arena
Document Arena (arena.ai)	Arena score (Elo scale)	1476	Independent	high	13 Sep 2026	Arena
Document Arena (arena.ai)	Votes	7411	Independent	high	13 Sep 2026	Arena
Text Arena (arena.ai)	Arena score (Elo scale)	1462	Independent	high	25 Sep 2026	Arena
Text Arena (arena.ai)	Votes	43340	Independent	high	25 Sep 2026	Arena
Vision Arena (arena.ai)	Arena score (Elo scale)	1265	Independent	high	28 Sep 2026	Arena
Vision Arena (arena.ai)	Votes	12722	Independent	high	28 Sep 2026	Arena

Other Claude Sonnet versions

Sources

Claude Sonnet 5: benchmarks, price and specs · supermarket.dev