Claude Sonnet 3.7
| Released | 2025-02-19 |
|---|---|
| Price per 1M | $3.00 in · $15.00 out |
| Overall score | 46.4 · rank 146 |
| Context window | 200,000 tokens |
| Provider | Anthropic |
| Weights | Proprietary |
Scores
| Dimension | Score | Rank | Benchmarks |
|---|---|---|---|
| Overall: Balanced | 46.4 | 146 | 24 |
| Overall: Coding | 46.5 | 141 | 12 |
| Overall: Reasoning | 45.7 | 155 | 10 |
| Overall: Agentic work | 46.9 | 138 | 12 |
| Overall: Frontend | 46.5 | 122 | 8 |
| Coding | 46.2 | 147 | 4 |
| Agents | 47.1 | 126 | 3 |
| Reasoning | 47.0 | 136 | 5 |
| Math | 45.6 | 145 | 3 |
| Knowledge | 42.8 | 128 | 2 |
| General | 48.8 | 122 | 3 |
| Multimodal | 47.7 | 82 | 4 |
| Value vs price | −6.6 | 164 |
Configurations
| Configuration | Balanced | Benchmarks |
|---|---|---|
| Unspecified setting | 44.9 ±0.9 | 23 |
| Thinking | 46.7 ±1.1 | 17 |
Evidence
Coding 46.2 ±1.9
| Benchmark | Result | Source | Date |
|---|---|---|---|
| Aider Polyglot | 64.9% | Epoch AI | 2026-11-11 |
| GSO | 3.8% | Epoch AI | 2026-11-11 |
| SWE-bench Verified (Epoch run) | 61.0% | Epoch AI | 2026-02-04 |
| LiveCodeBench (Vals run) | 60.4% | Vals AI | 2026-09-01 |
Agents 47.1 ±3.4
| Benchmark | Result | Source | Date |
|---|---|---|---|
| DeepResearch Bench | 43.6% | Epoch AI | 2026-11-11 |
| METR 50% Time Horizon | $60 | METR | 2026-11-11 |
| OSWorld-Verified | 35.8% | OSWorld-Verified | 2025-07-28 |
| OSWorld (v1) | 35.8% | Epoch AI | 2026-11-11 |
| METR task suite | 60.0% | Epoch AI | 2026-11-11 |
Reasoning 47.0 ±2.3
| Benchmark | Result | Source | Date |
|---|---|---|---|
| ARC-AGI-2 | 0.9% | Epoch AI | 2026-11-11 |
| ARC-AGI-1 | 28.6% | Epoch AI | 2026-11-11 |
| EnigmaEval | 4.2% | Epoch AI | 2026-11-11 |
| GPQA Diamond | 79.7% | Epoch AI | 2025-05-26 |
| GPQA Diamond (Vals run) | 75.3% | Vals AI | 2026-09-01 |
| SimpleBench | 46.4% | SimpleBench | 2026-11-11 |
| SimpleBench | 46.4% | Epoch AI | 2026-11-11 |
Math 45.6 ±2.0
| Benchmark | Result | Source | Date |
|---|---|---|---|
| FrontierMath Tiers 1-3 (2025 set) | 4.1% | Epoch AI | 2025-03-06 |
| MATH Level 5 | 91.2% | Epoch AI | 2025-03-13 |
| OTIS Mock AIME 2024-2025 | 57.8% | Epoch AI | 2025-03-13 |
Knowledge 42.8 ±1.3
| Benchmark | Result | Source | Date |
|---|---|---|---|
| Humanity's Last Exam | 8.0% | Epoch AI | 2026-11-11 |
| MMLU-Pro (Vals run) | 82.7% | Vals AI | 2026-09-01 |
General 48.8 ±3.2
| Benchmark | Result | Source | Date |
|---|---|---|---|
| LMArena Text | 1,388 | LMArena | 2026-11-09 |
| EQ-Bench Creative Writing v3 | 1,412 | EQ-Bench | 2026-11-11 |
| EQ-Bench Longform Creative Writing | 57.7% | EQ-Bench | 2026-11-11 |
Multimodal 47.7 ±2.8
| Benchmark | Result | Source | Date |
|---|---|---|---|
| LMArena Vision | 1,193 | LMArena | 2026-11-09 |
| MMMU-Pro (Vals run) | 75.1% | Vals AI | 2026-09-01 |
| VPCT | 39.0% | Epoch AI | 2026-11-11 |
| GeoBench | 68.0% | Epoch AI | 2026-11-11 |
Results not used in the score
| Benchmark | Result | Source | Why not used |
|---|---|---|---|
| ARC-AGI-1 | 13.6% | Epoch AI | A stronger configuration result was used |
| ARC-AGI-2 | 0.0% | Epoch AI | A stronger configuration result was used |
| Aider Polyglot | 60.4% | Epoch AI | A stronger configuration result was used |
| FrontierMath Tiers 1-3 (2025 set) | 3.1% | Epoch AI | A stronger configuration result was used |
| GPQA Diamond | 66.0% | Epoch AI | A stronger configuration result was used |
| GPQA Diamond (Vals run) | 67.4% | Vals AI | A stronger configuration result was used |
| GeoBench | 65.0% | Epoch AI | A stronger configuration result was used |
| LMArena Text | 1,372 | LMArena | A stronger configuration result was used |
| LMArena Vision | 1,175 | LMArena | A stronger configuration result was used |
| LiveCodeBench (Vals run) | 56.7% | Vals AI | A stronger configuration result was used |
| MATH Level 5 | 68.2% | Epoch AI | A stronger configuration result was used |
| METR task suite | 55.8% | Epoch AI | A stronger configuration result was used |
| MMLU-Pro (Vals run) | 80.7% | Vals AI | A stronger configuration result was used |
| MMMU-Pro (Vals run) | 71.5% | Vals AI | A stronger configuration result was used |
| OSWorld (v1) | 35.6% | Epoch AI | A stronger configuration result was used |
| OSWorld (v1) | 27.1% | Epoch AI | A stronger configuration result was used |
| OSWorld-Verified | 27.1% | OSWorld-Verified | A stronger configuration result was used |
| OSWorld-Verified | 35.6% | OSWorld-Verified | A stronger configuration result was used |
| OTIS Mock AIME 2024-2025 | 21.9% | Epoch AI | A stronger configuration result was used |
| SimpleBench | 46.4% | Epoch AI | A stronger configuration result was used |
| SimpleBench | 44.9% | SimpleBench | A stronger configuration result was used |
| VPCT | 35.0% | Epoch AI | A stronger configuration result was used |
Names matched to this model
claude-sonnet-3-7, claude-sonnet-3-7-20250219