[whatmodel.sh](../../README.md) / Claude Sonnet 5
# Claude Sonnet 5
Anthropic · released 2026-06 · closed weights
**score 73.4** ±0.5 · #46 of 732 · from 42 benchmark results across 11 sources
Score within the [LMArena (text)](https://whatmodel.sh/?set=lmarena-text&score=set) comparison set only.
## What it's good at
● marks Claude Sonnet 5 among every model on that task or test, worst to best.
task vs the others rank counts tests
agent tasks ▁▁▃▆█▄▅▅▄●▂▁ #36 of 313 38% 14
coding ▁▁▁▃▅▇█▇▅●▃▁ #44 of 469 23% 7
chat & writing ▁▁▁▃▆▄▆█▆▇●▃ #42 of 402* 2.6% 1
math ▁▁▃▅█▆▄█●▂▁▁ #37 of 431 23% 4
vision & documents ▁▁▂▂▁▃▄▂▆▇●▄ #34 of 153 4.6% 4
web search ▁▁▁▅▄●▃▄·▁·▁ #33 of 84 3.8% 3
agent tasks · #36 of 313 · 38% of its score
Beats [GPT-5.6 Terra](gpt-5-6-terra.md), [GPT-5.6 Luna](gpt-5-6-luna.md), [Grok 4.6](grok-4-6.md) here, though they rank higher overall. Trails [Grok 4.7](grok-4-7.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
SWE-bench Pro (split unspecified)… 63.2% ▄●·▄·▄██···▄ beats 1/9 very hard · avg 7.4% 6.9% sharp · at its level
Arena Agent (October 2026 aggrega… 0.0294 IPS ▁▁▁▃▃▂●▁▂▂▁▁ beats 42/59 avg -0.379 IPS 6.6% sharp
Terminal-Bench 4.0 12.4% ●▅▅▅▅·▅▃▃▅▅▅ beats 1/20 very hard · avg 0.7% 4.8% sharp · several runs
SWE-bench Multilingual (maker) 78.3% ▃··●·▅▃▃█··▃ beats 1/9 hard · avg 31% 3.5% typical
SWE-bench Verified (maker) 85.2% ▁▁▁▁▁▁▂▄█●▁▁ beats 134/138 hard · avg 28% 3.2% sharp
Terminal-Bench 2.1 (maker) 80.4% ▁·▁·▁▁▄▃▆█●█ beats 44/63 hard · avg 19% 2.0% sharp
OSWorld-Verified (maker) 81.2% █▃·▂▅▂▃▂▅▃●▆ beats 22/27 hard · avg 28% 2.0% typical
GameLogicBench solve rate (72 God… 25.0% █▄·●·▄·██▄·▄ beats 4/11 very hard · avg 5.1% 1.7% at its level
Humanity's Last Exam with tools (… 54.9% ▁······▁·●·█ beats 2/10 very hard · avg 14% 1.7% at its level
E2E-SWE (build a working codebase… 43.1% ▃··▅··▃●▃▅▃▅ beats 5/12 very hard · avg 3.7% 1.5% sharp · at its level
Toolathlon (Tool Decathlon) (make… 54.3% █···●··█·███ beats 1/5 hard · avg 16% 1.4% at its level
BrowseComp with context managemen… 84.7% ▄·▄▂·█▂·●▄·▂ beats 12/15 medium · avg 47% 1.2% typical
Toolathlon-Verified (maker) 74.7% ●·····█···▃▃ beats 0/5 medium · avg 37% 0.8% blunt · at its level
KlinikeBench strict pass@1 (333 i… 13.8% ▁▁·▁▁▁▁▅▃●▅█ beats 12/27 very hard · avg 2.4% 0.7% typical
coding · #44 of 469 · 23% of its score
Beats [GPT-5.5](gpt-5-5.md), [GPT-5.6 Terra](gpt-5-6-terra.md), [GPT-5.4](gpt-5-4.md) here, though they rank higher overall. Trails [Qwen3.8 27B](qwen3-8-27b.md), [GLM-5.2](glm-5-2.md), [GLM-5.3-Flash](glm-5-3-flash.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
SWE-bench Pro (split unspecified)… 63.2% ▄●·▄·▄██···▄ beats 1/9 very hard · avg 7.4% 6.9% sharp · at its level
Arena WebDev (Code Arena) 1540.1 ▁▁▃▂▅█▃●▅▅▁▂ beats 84/129 avg 981 4.6% sharp
SWE-bench Multilingual (maker) 78.3% ▃··●·▅▃▃█··▃ beats 1/9 hard · avg 31% 3.5% typical
SWE-bench Verified (maker) 85.2% ▁▁▁▁▁▁▂▄█●▁▁ beats 134/138 hard · avg 28% 3.2% sharp
Luau Leetcode fully solved (Torpe… 68.3% ▁▁▁▁·▁▂▃●▃█▇ beats 21/49 hard · avg 17% 1.9% sharp · at its level
GameLogicBench solve rate (72 God… 25.0% █▄·●·▄·██▄·▄ beats 4/11 very hard · avg 5.1% 1.7% at its level
E2E-SWE (build a working codebase… 43.1% ▃··▅··▃●▃▅▃▅ beats 5/12 very hard · avg 3.7% 1.5% sharp · at its level
chat & writing · #42 of 402 · 2.6% of its score
Beats [GPT-6 Sol](gpt-6-sol.md), [MiMo-V2.6-Flash (309B-A15B)](mimo-v2-6-flash.md), [Claude Haiku 5.5](claude-haiku-5-5.md) here, though they rank higher overall. Trails [MiMo-V2.5-Pro (1T-A42B)](mimo-v2-5-pro.md), [GLM-5.1](glm-5-1.md), [Grok 4.20 (0309 reasoning)](grok-4-20.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
LMArena (text) 1461.7 ▁▁▁▂▃▃▃▆▆▇●▃ beats 336/385 avg 1260 2.6% blunt
math · #37 of 431 · 23% of its score
Beats [DeepSeek V4 Pro (0813)](deepseek-v4-pro.md), [Claude Opus 4.7](claude-opus-4-7.md), [Grok 4.5](grok-4-5.md) here, though they rank higher overall. Trails [GPT-5.2](gpt-5-2.md), [Qwen3.7 Max](qwen3-7-max.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
FrontierMath tiers 1–3 v2 65.6% ▂··▁▁▃▄▄●▆▄▃ beats 52/86 very hard · avg 1.2% 12% sharp · at its level
FrontierMath tier 4 v2 29.3% ▁▂▁▂●▅▂▃▂▁▁▁ beats 34/67 very hard · avg 0.1% 10% sharp · at its level
USAMO 2026 (maker) 79.5% █·█●······██ beats 2/4 hard · avg 19% 0.7% sharp · noisy
OTIS mock AIME 2024–2025 80.0% ▃▃▄▃▃▃▇●█▃▂▅ beats 114/197 very hard · avg 7.2% 0.4% sharp · maxed out · noisy
vision & documents · #34 of 153 · 4.6% of its score
Beats [Claude Sonnet 5.5](claude-sonnet-5-5.md), [GPT-6 Sol](gpt-6-sol.md), [MiMo-V2.6-Pro](mimo-v2-6-pro.md) here, though they rank higher overall. Trails [Qwen3.6 Flash](qwen3-6-flash.md), [Qwen3.6 Plus](qwen3-6-plus.md), [Qwen3.7 Plus](qwen3-7-plus.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
Arena vision (style controlled) 1261.9 ▁▁▂▂▂▂▃▄▅▅●▇ beats 116/152 avg 1061 1.7% blunt
CharXiv (reasoning questions) (ma… 77.0% ███··●·██··█ beats 3/6 hard · avg 28% 1.7% typical
Arena document (style controlled) 1475.7 ▁▁▃▃▄▃▇█▅●▄▅ beats 26/37 avg 1331 0.9% blunt
SpaceCast-Bench predictive spatia… 38.3% ▃·▅▃●▃··▅··█ beats 5/11 very hard · avg 33% 0.3% at its level · noisy
web search · #33 of 84 · 3.8% of its score
Beats [Gemini 3.1 Pro (Preview)](gemini-3-1-pro.md), [Claude Opus 4.7](claude-opus-4-7.md) here, though they rank higher overall. Trails [GPT-4o mini (2024-07-18)](gpt-4o-mini-2024-07-18.md), [Mistral Small 3.1 24B (2503)](mistral-small-3-1.md), [GPT-4o (2024-08-06)](gpt-4o-2024-08-06.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
Humanity's Last Exam with tools (… 54.9% ▁······▁·●·█ beats 2/10 very hard · avg 14% 1.7% at its level
BrowseComp with context managemen… 84.7% ▄·▄▂·█▂·●▄·▂ beats 12/15 medium · avg 47% 1.2% typical
Arena search systems (style contr… 1193.2 ▁·▁▁▂▂▂●▆▃▂▂ beats 16/36 avg 1069 0.9% blunt
other tests (15)
test result vs the others in its run difficulty counts why
Mystery Game Puzzles (Epoch) 35.0% █··▂▃▇▄●▂▄▂▁ beats 56/77 very hard · avg 9.4% 3.6% sharp · noisy
KMMLU-HARD (Daekeun Kim, zero-sho… 77.4% █···▁▁▁▂▃·●▂ beats 20/23 very hard · avg 29% 3.3% sharp · at its level
Humanity's Last Exam (maker) 43.1% ▂·····▂·●▂█▄ beats 2/10 very hard · avg 5.9% 2.9% at its level
Chess Puzzles (Epoch) 16.0% █▁▁▁▁▂●▃▃▂▁▁ beats 70/137 very hard · avg 5.8% 2.3% sharp · at its level · noisy
GPQA Diamond 80.3% ▁▁▁▁▂▄▆▄▅●█▅ beats 129/211 hard · avg 39% 1.9% sharp
CLIcK Korean (Daekeun Kim) 92.1% ▆▄▄▄█▄▂·▄●·▄ beats 20/23 medium · avg 56% 1.6% typical
Global-MMLU (maker) 89.2% ▄●····▄▄··▄█ beats 1/6 medium · avg 64% 1.5% typical
KoBALT-700 (Daekeun Kim) 76.1% ▄●█·····▄··▄ beats 1/5 hard · avg 39% 1.2% at its level
HAE-RAE 1.0 (Daekeun Kim) 92.0% ▄▄▄▂▂█▂▄▂●▂▄ beats 18/22 medium · avg 61% 1.2% maxed out
KMMLU-Pro (Daekeun Kim) 85.8% ▄·▄●·····▄·▄ beats 3/5 medium · avg 43% 1.1% typical
SimpleQA Verified 33.7% ▁▁▃▂▃▃●▆█▄▂▅ beats 26/81 very hard · avg 8.0% 1.0% at its level
Icelandic LLM Leaderboard average… 55.2% ▁▁▃▂▂▄▁▃▄●▄█ beats 24/38 very hard · avg 14% 1.0% at its level
MuSR Korean (Daekeun Kim) 81.6% █·····█●█·██ beats 2/5 medium · avg 45% 0.7% typical
HealthBench (maker) 59.2% ●·▄·····▄▄▄▄ beats 1/6 hard · avg 30% 0.7% blunt · at its level
RobloxQA v2.0 multiple choice (To… 94.0% ▁▁▃▁█▄●▃▆▂▂▂ beats 26/49 easy · avg 77% 0.7% maxed out
source results share of the score
Anthropic system cards and model announ… 18 33% ▇▇▇▇
Epoch AI 7 31% ▇▇▇▇
LMArena leaderboard dataset 6 17% ▇▇
Daekeun Kim Korean LLM benchmark runs 6 9% ▇
Roblox LLM Leaderboard 2 3% ▇
Terminal-Bench 4.0 official leaderboard… 1 2% ▇
GameLogicBench: Evaluating Coding Agent… 1 2% ▇
E2E-SWE: Benchmarking LLMs on Building … 1 2% ▇
Miðeind Icelandic LLM Leaderboard 1 1% ▇
KlinikeBench: Evaluating Language Model… 1 1% ▇
SpaceCast-Bench: Evaluating Predictive … 1 0% ▇
prior: pull toward the middle 0% ▇
weighted average = 73.4
Licences: Epoch AI ([CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)) · LMArena leaderboard dataset ([CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)) · Daekeun Kim Korean LLM benchmark runs ([MIT](../../licenses/korean-personal-benchmark-runs-MIT.txt)) · Roblox LLM Leaderboard ([MIT](../../licenses/roblox-llm-leaderboard-MIT.txt)) · Terminal-Bench 4.0 official leaderboard submissions ([Apache-2.0](../../licenses/terminal-bench-leaderboard-submissions-Apache-2.0.txt)) · GameLogicBench: Evaluating Coding Agents on Runtime Game Logic with Tick-Level State Assertions ([CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)) · E2E-SWE: Benchmarking LLMs on Building Working Codebases from Scratch ([CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)) · Miðeind Icelandic LLM Leaderboard ([CC BY 4.0](../../licenses/mideind-icelandic-llm-leaderboard-CC-BY-4.0.txt)) · KlinikeBench: Evaluating Language Models Beyond Diagnostic Accuracy ([CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)) · SpaceCast-Bench: Evaluating Predictive Spatial Reasoning in Vision-Language Models ([CC BY 4.0](https://creativecommons.org/licenses/by/4.0/))
## How to run it
### Through an API
api (Anthropic) · $4.00/M blended · $2 in / $10 out per 1M tokens · [docs.anthropic.com](https://docs.anthropic.com/en/docs/about-claude/models)
[price source](https://github.com/anomalyco/models.dev/blob/a668908cfe7073faa0363699486fc41a3ff52412/providers/anthropic/models/claude-sonnet-5.toml): Standard base token rate from models.dev. Provider-specific tiers, long-context surcharges, caching and promotions may differ. · [context source](https://github.com/anomalyco/models.dev/blob/a668908cfe7073faa0363699486fc41a3ff52412/models/anthropic/claude-sonnet-5.toml)
[interactive version](https://whatmodel.sh/model/claude-sonnet-5) · [JSON](https://whatmodel.sh/api?format=json&all=1)