AnthropicClaude 3.7 Sonnet
Claude 3.7 Sonnet is an Anthropic proprietary reasoning model, released 24 Feb 2025. Its best configuration (thinking reasoning effort) ranks #352 of 785 on the BenchLeader Index at 51.0 ±2.8, in the upper half. The ± is the point: 301 other configurations score within that range, so they and this one cannot be told apart on quality alone — price and speed are what separate them. It scores highest in human preference (56) and lowest in instruction following (45). It has been measured at 3 reasoning-effort settings; this summary describes the best-scoring one, and the tabs above switch between them. Last measured 22 Oct 2026.
- Blended price
- –
- Output speed
- –
- First answer
- –
- Context
- 200k
- Full answer
- –
- Cost per run
- –
- Released
- 24 Feb 2025
- Overall index51
- Reasoning47
- Coding48
- Agents & tools52
- Maths53
- Knowledge55
- Instruction following45
- Human preference56
- Multimodal54
- Composite49
Versions
Anthropic has shipped 10 models under this name. Each is ranked on its own results; a newer version often has fewer results so far, which holds its index nearer the average until more arrive.
| Model | Released | Index | Rank |
|---|---|---|---|
| Claude Sonnet 5.5max | 28 Sept 2026 | 69.4 | #10 |
| Claude Sonnet 5high | 30 Jun 2026 | 60.1 | #116 |
| Claude Sonnet 4.6 | 17 Feb 2026 | 57.1 | #185 |
| Claude Sonnet 4.5high | 29 Sept 2025 | 58.1 | #161 |
| Claude Sonnet 4thinking | 22 May 2025 | 53.2 | #293 |
| Claude 3.7 Sonnetthinkingthis page | 24 Feb 2025 | 51.0 | #352 |
| Claude 3.5 Sonnet | 22 Oct 2024 | 46.3 | #490 |
| Claude 3 Sonnet | 29 Feb 2024 | 38.0 | #721 |
| Claude 37 Sonnetthinking | – | – | – |
| Claude Sonnet 5.0max | – | – | – |
Reasoning-effort configurations
The same model behaves differently depending on how much it is allowed to think. Each row is one setting, scored only on the benchmarks that were run at that setting. “Not stated” collects results from publishers that did not say which setting they used; for a reasoning model that is usually its thinking mode, but we do not assume it. Pick a setting here or at the top of the page to see its price, speed and category scores.
| Effort | Index | Rank | Speed | First answer | Chat reply cost | Categories |
|---|---|---|---|---|---|---|
| no reasoning | 49.5 | #394 | – | – | $0.0057 | Agents & tools 52 · Coding 59 · Composite 47 · Instruction following 46 · Knowledge 57 · Long context 50 · Multimodal 43 · Reasoning 41 |
| thinkingbest | 51.0 | #352 | – | – | – | Agents & tools 52 · Coding 48 · Composite 49 · Human preference 56 · Instruction following 45 · Knowledge 55 · Maths 53 · Multimodal 54 · Reasoning 47 |
| not stated | 50.0 | #382 | – | – | – | Agents & tools 44 · Coding 48 · Human preference 54 · Instruction following 54 · Knowledge 52 · Long context 48 · Maths 50 · Multimodal 51 · Reasoning 49 |
How its index has moved
8 Sept 2026 to 21 Oct 2026- thinking
- not stated
- no reasoning
The index is recomputed from scratch every day, so a line moves when a new benchmark result lands, when a publisher revises a score, or when the models it is normalised against change. Early movement usually means the score is still settling.
Benchmark results
See all LLM benchmarks. One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.
Reasoning
| Benchmark | no reasoning | thinking | not stated | Source | Trend |
|---|---|---|---|---|---|
| GPQA Diamondeffort not stated | – | – | 79.7%#127 | Epoch AI Benchmarking Hub | |
| Humanity's Last Exameffort not stated | – | – | 8.0%#35 | Scale AI / CAIS | |
| SimpleBencheffort not stated | – | – | 46.4%#51 | SimpleBench | |
| Arena Hard Prompts | – | 1418#152 | 1399#174 | Arena (LMArena) | |
| GPQA Diamond (AA)not in index | 65.6%#341 | 77.2%#223 | – | Artificial Analysis | |
| Humanity's Last Exam (AA)not in index | 4.2%#535 | 9.7%#331 | – | Artificial Analysis | |
| GPQA Diamond (Vals)not in index | – | 75.3%#77 | 67.4%#96 | Vals AI | |
| GPQA Diamond (HELM)not in indexeffort not stated | – | – | 60.8%#26 | HELM Capabilities | |
| CritPt | 0.0%#315 | 0.9%#237 | – | Artificial Analysis | |
| ForecastBenchnot in indexeffort not stated | – | – | 64.0%#37 | ForecastBench |
Coding
| Benchmark | no reasoning | thinking | not stated | Source | Trend |
|---|---|---|---|---|---|
| SWE-bench Verified (Epoch)effort not stated | – | – | 61.0%#29 | Epoch AI Benchmarking Hub | |
| GSO-Bencheffort not stated | – | – | 3.8%#30 | GSO-Bench | |
| Arena Coding | – | 1452#143 | 1431#174 | Arena (LMArena) | |
| LiveCodeBench | – | 60.4%#101 | 56.7%#106 | Vals AI | |
| Aider Polyglot | 60.4%#14 | – | 64.9%#10 | Aider polyglot leaderboard | |
| SWE-bench Verified (bash only)effort not stated | – | – | 52.8%#33 | SWE-bench | |
| SWE-bench Verified (any scaffold)not in indexeffort not stated | – | – | 66.4%#18 | SWE-bench | |
| CadEvalnot in indexeffort not stated | – | – | 54.0%#5 | Epoch AI Benchmarking Hub |
Agents & tools
| Benchmark | no reasoning | thinking | not stated | Source | Trend |
|---|---|---|---|---|---|
| Cybencheffort not stated | – | – | 20.0%#11 | Cybench | |
| Terminal-Bench Hard | 21.2%#158 | 21.2%#158 | – | Artificial Analysis | |
| τ²-Bench Telecom (AA)not in index | 50.0%#197 | 54.7%#186 | – | Artificial Analysis | |
| DeepResearch Benchnot in indexeffort not stated | – | – | 43.6%#32 | DeepResearch Bench | |
| TheAgentCompanynot in indexeffort not stated | – | – | 30.9%#4 | TheAgentCompany | |
| METR Time Horizonsnot in indexeffort not stated | – | – | 60.0%#20 | METR | |
| BrowseCompnot in indexeffort not stated | – | – | 0.9%#21 | OpenAI |
Maths
| Benchmark | thinking | not stated | Source | Trend |
|---|---|---|---|---|
| OTIS Mock AIMEeffort not stated | – | 57.8%#167 | Epoch AI Benchmarking Hub | |
| MATH Level 5effort not stated | – | 91.2%#19 | Epoch AI Benchmarking Hub | |
| AIME (Vals) | 44.6%#61 | 22.3%#76 | Vals AI | |
| MGSM | 93.0%#12 | 92.4%#19 | Vals AI | |
| Omni-MATH (HELM)effort not stated | – | 33.0%#38 | HELM Capabilities | |
| MATH 500not in index | 91.6%#21 | 76.8%#41 | Vals AI | |
| Arena Maths | 1386#168 | 1363#187 | Arena (LMArena) |
Knowledge
| Benchmark | no reasoning | thinking | not stated | Source | Trend |
|---|---|---|---|---|---|
| AA-Omniscience | -9.7#188 | -0.7#141 | – | Artificial Analysis | |
| MMLU-Pro | – | 82.7%#73 | 80.7%#81 | Vals AI | |
| LegalBench | – | 82.5%#67 | 80.0%#88 | Vals AI | |
| CorpFin | – | 60.4%#68 | – | Vals AI | |
| TaxEval | – | 74.0%#41 | 72.4%#61 | Vals AI | |
| MedQA | – | 90.2%#46 | – | Vals AI | |
| MultiNRC | – | 27.8%#26 | – | Scale AI SEAL | |
| MMLU-Pro (HELM)not in indexeffort not stated | – | – | 78.4%#24 | HELM Capabilities | |
| AA-Omniscience: accuracynot in index | 28.2%#200 | 28.0%#201 | – | Artificial Analysis | |
| AA-Omniscience: non-hallucinationnot in index | 47.2%#135 | 60.0%#77 | – | Artificial Analysis |
Instruction following
| Benchmark | no reasoning | thinking | not stated | Source | Trend |
|---|---|---|---|---|---|
| IFBench | 44.0%#218 | 48.3%#184 | – | Artificial Analysis | |
| MultiChallenge | – | 51.6%#25 | – | Scale AI SEAL | |
| IFEval (HELM)effort not stated | – | – | 83.4%#32 | HELM Capabilities | |
| WildBench (HELM)effort not stated | – | – | 81.4%#26 | HELM Capabilities | |
| Arena Instruction Followingnot in index | – | 1411#122 | 1385#155 | Arena (LMArena) |
Human preference
| Benchmark | thinking | not stated | Source | Trend |
|---|---|---|---|---|
| Arena Text | 1388#164 | 1373#182 | Arena (LMArena) | |
| Arena Creative Writingnot in index | 1393#115 | 1377#138 | Arena (LMArena) | |
| Arena Multi-turnnot in index | 1400#153 | 1399#154 | Arena (LMArena) | |
| Arena Longer Queriesnot in index | 1429#121 | 1412#144 | Arena (LMArena) | |
| Lech Mazur Creative Writingnot in indexeffort not stated | – | 81.1%#15 | Lech Mazur |
Multimodal
| Benchmark | no reasoning | thinking | not stated | Source | Trend |
|---|---|---|---|---|---|
| Arena Vision | – | 1169#107 | 1151#117 | Arena (LMArena) | |
| MMMU-Pro | 60.1%#212 | – | – | Artificial Analysis | |
| VISTA | – | 48.2%#16 | 43.0%#35 | Scale AI SEAL | |
| MMMU (validation)effort not stated | – | – | 75.0%#14 | MMMU | |
| MMMU-Pro (Vals)not in index | – | 75.1%#54 | 71.5%#66 | Vals AI | |
| MortgageTaxnot in index | – | 66.8%#31 | 68.7%#12 | Vals AI | |
| GeoBenchnot in indexeffort not stated | – | – | 68.0%#13 | GeoBench | |
| VPCTnot in indexeffort not stated | – | – | 39.0%#19 | Epoch AI Benchmarking Hub | |
| SpatialViz-Benchnot in indexeffort not stated | – | – | 33.9%#5 | Epoch AI Benchmarking Hub |
Long context
| Benchmark | no reasoning | not stated | Source | Trend |
|---|---|---|---|---|
| Fiction.LiveBench 120keffort not stated | – | 53.1%#20 | Fiction.live | |
| AA-LCR | 51.7%#319 | – | Artificial Analysis |
Composite
| Benchmark | no reasoning | thinking | not stated | Source | Trend |
|---|---|---|---|---|---|
| Epoch Capabilities Indexnot in indexeffort not stated | – | – | 141.0#109 | Epoch AI Benchmarking Hub | |
| AA Intelligence Index v4.3.2 | 15.3#285 | 17.7#261 | – | Artificial Analysis | |
| HELM Capabilities meannot in indexeffort not stated | – | – | 67.4%#28 | HELM Capabilities |
Safety & honesty
| Benchmark | thinking | not stated | Source | Trend |
|---|---|---|---|---|
| FORTRESSnot in indexeffort not stated | – | 38.0%#21 | Scale AI SEAL | |
| MASKnot in index | 82.1%#23 | 72.3%#30 | Scale AI SEAL |
More from Anthropic
See also
Questions people ask
- What are Claude 3.7 Sonnet's benchmark scores?
- As of 25 Oct 2026, Claude 3.7 Sonnet (thinking effort) scores 51.0 ±2.8 on the BenchLeader Index, strongest in human preference (56) and weakest in instruction following (45). That rests on 21 independent benchmark results, and every score on this page links to its publisher.
- How does Claude 3.7 Sonnet rank?
- Claude 3.7 Sonnet (thinking effort) is #352 of 785 ranked configurations on 25 Oct 2026. 301 other configurations score within its margin, so the exact position matters less than the band. The leader is Claude Opus 5.5 (max effort) at 72.6.
- When was Claude 3.7 Sonnet released?
- Anthropic released Claude 3.7 Sonnet on 24 Feb 2025. BenchLeader has results for it at 3 reasoning-effort settings.
Data as of 25 Oct 2026. Compare these configurations.
Cite as: BenchLeader, “Claude 3.7 Sonnet: benchmarks, pricing, speed and rank”, data as of 25 Oct 2026, https://www.benchleader.com/models/claude-3-7-sonnet.