Model comparison
Claude 3.7 Sonnet vs Claude Opus 4.6
Claude Opus 4.6 is the stronger model overall, scoring 58.3 to 39.4 on the Noometry Index.
Last verified . 35 shared benchmarks.
Summary
- They share 35 benchmarks with published results for both. Claude 3.7 Sonnet scores higher in 1 category and Claude Opus 4.6 in 9 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Claude Opus 4.6 leads 57.7 to 18.7.
- The biggest single-benchmark swing is Cybench: 20% for Claude 3.7 Sonnet and 93% for Claude Opus 4.6.
Side by side
| Claude 3.7 Sonnet | Claude Opus 4.6 | |
|---|---|---|
| Provider | Anthropic | Anthropic |
| Noometry Index | 39.4 | 58.3 |
| Released | 2025-02-24 | 2026-02-04 |
| Weights | Proprietary | Proprietary |
| Context window | — | 1M |
| Max output | — | 128K |
| Input $ / M tokens | — | $5 |
| Output $ / M tokens | — | $25 |
| Results tracked | 58 | 68 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Claude Opus 4.6 leads
Claude 3.7 Sonnet: 41.1 (#120), Claude Opus 4.6: 57.7 (#15)
| Benchmark | Claude 3.7 Sonnet | Claude Opus 4.6 |
|---|---|---|
| SWE-bench Verified | 61% | 78.7% |
| SWE-bench Verified (bash only) | 52.8% | 75.6% |
| GSO | 3.8% | 41.2% |
| LMArena Coding | 1361 | 1535 |
| FrontierCode | — | 26.6% |
| Aider Polyglot | 64.9% | — |
| LMArena WebDev | — | 1546 |
| SWE-bench Multilingual | — | 72% |
| WeirdML | — | 78% |
| LiveBench Coding | 74.5% | — |
| CadEval | 54% | — |
| ALE-Bench | — | 996.5 |
| AlgoTune | — | 1.47 |
Agentic & Tool Use Claude Opus 4.6 leads
Claude 3.7 Sonnet: 33.9 (#46), Claude Opus 4.6: 51.0 (#4)
| Benchmark | Claude 3.7 Sonnet | Claude Opus 4.6 |
|---|---|---|
| Cybench | 20% | 93% |
| DeepResearch Bench | 43.6% | 55.3% |
| METR Time Horizons | 60% | 78.9% |
| Terminal-Bench | — | 79.8% |
| APEX-Agents | — | 46.3% |
| Remote Labor Index | — | 4.2% |
| TheAgentCompany | 30.9% | — |
| τ²-bench Banking | — | 27.3% |
| OSWorld | 35.8% | — |
| GBAEval | — | 44.1% |
| LMArena Search | — | 1253 |
| Vending-Bench 2 | — | 8,018 |
Reasoning Claude Opus 4.6 leads
Claude 3.7 Sonnet: 18.7 (#252), Claude Opus 4.6: 57.7 (#21)
| Benchmark | Claude 3.7 Sonnet | Claude Opus 4.6 |
|---|---|---|
| ARC-AGI-2 | 0.9% | 69.2% |
| SimpleBench | 46.4% | 67.6% |
| ARC-AGI-1 | 28.6% | 94% |
| EnigmaEval | 4.2% | 7.6% |
| LMArena Hard Prompts | 1333 | 1527 |
| Epoch Capabilities Index | 141.16 | 155.36 |
| ForecastBench | 61.8 | 60 |
| Kagi LLM Benchmark | — | 83.6% |
| NYT Connections (extended) | — | 92.1% |
| Chess Puzzles | — | 17% |
| Thematic Generalization | — | 80.6% |
| EBR-Bench | — | 12.7% |
| LiveBench Reasoning | 87.8% | — |
| Mystery Game Puzzles | — | 25% |
| DTBench | — | 91.2% |
| LiveBench Data Analysis | 74% | — |
| LMCA | — | 55.8% |
| LiveBench | 76.1% | — |
Math Claude Opus 4.6 leads
Claude 3.7 Sonnet: 35.7 (#137), Claude Opus 4.6: 62.7 (#25)
| Benchmark | Claude 3.7 Sonnet | Claude Opus 4.6 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 57.8% | 94.4% |
| LMArena Math | 1336 | 1517 |
| FrontierMath (Feb 2025 set) | 4.1% | 40.7% |
| FrontierMath (Tiers 1-3) | — | 66% |
| FrontierMath Tier 4 | — | 26.8% |
| MathArena Final-Answer Competitions | — | 78.5% |
| ProofBench | — | 50% |
| Omni-MATH | 33% | — |
| LiveBench Math | 79% | — |
| MATH Level 5 | 91.2% | — |
| FrontierMath Tier 4 (v1) | — | 22.9% |
Knowledge Claude Opus 4.6 leads
Claude 3.7 Sonnet: 39.7 (#117), Claude Opus 4.6: 61.8 (#23)
| Benchmark | Claude 3.7 Sonnet | Claude Opus 4.6 |
|---|---|---|
| GPQA Diamond | 79.7% | 90.5% |
| Humanity's Last Exam | 8% | 34.4% |
| LMArena Expert | 1321 | 1546 |
| SimpleQA Verified | — | 47% |
| MMLU-Pro | 78.4% | — |
| Confabulations | 14.7% | — |
| Vectara Hallucination Rate | — | 12.2% |
| GPQA (HELM) | 60.8% | — |
Multimodal Claude Opus 4.6 leads
Claude 3.7 Sonnet: 33.1 (#90), Claude Opus 4.6: 37.0 (#67)
| Benchmark | Claude 3.7 Sonnet | Claude Opus 4.6 |
|---|---|---|
| LMArena Vision | 1169 | 1316 |
| GeoBench | 68% | — |
| VPCT | 39% | — |
| Furniture Assembly | — | 28.3% |
| LMArena Document | — | 1507 |
| SpatialViz-Bench | 33.9% | — |
Multilingual Claude Opus 4.6 leads
Claude 3.7 Sonnet: 44.3 (#174), Claude Opus 4.6: 58.2 (#4)
| Benchmark | Claude 3.7 Sonnet | Claude Opus 4.6 |
|---|---|---|
| LMArena Non-English | 1296 | 1490 |
| LMArena Chinese | 1299 | 1551 |
| LMArena French | 1302 | 1513 |
| LMArena German | 1301 | 1504 |
| LMArena Japanese | 1268 | 1484 |
| LMArena Korean | 1249 | 1465 |
| LMArena Russian | 1311 | 1497 |
| LMArena Spanish | 1299 | 1512 |
Instruction Following Claude Opus 4.6 leads
Claude 3.7 Sonnet: 73.1 (#120), Claude Opus 4.6: 79.7 (#3)
| Benchmark | Claude 3.7 Sonnet | Claude Opus 4.6 |
|---|---|---|
| LMArena Instruction Following | 1352 | 1523 |
| LiveBench Instruction Following | 81.3% | — |
| IFEval | 83.4% | — |
Long Context Claude 3.7 Sonnet leads
Claude 3.7 Sonnet: 50.2 (#10), Claude Opus 4.6: 48.1 (#13)
| Benchmark | Claude 3.7 Sonnet | Claude Opus 4.6 |
|---|---|---|
| LMArena Longer Query | 1373 | 1520 |
| Fiction.LiveBench | 83.3% | — |
| CL-bench | — | 20.7% |
| CL-bench Life | — | 17% |
Writing & Preference Claude Opus 4.6 leads
Claude 3.7 Sonnet: 54.6 (#145), Claude Opus 4.6: 73.7 (#9)
| Benchmark | Claude 3.7 Sonnet | Claude Opus 4.6 |
|---|---|---|
| LMArena Text | 1314 | 1504 |
| LMArena Creative Writing | 1332 | 1506 |
| EQ-Bench Creative Writing | 1412 | 1809 |
| LMArena Multi-Turn | 1339 | 1514 |
| Short-Story Creative Writing | 81.1% | — |
| WildBench | 81.4% | — |
| EQ-Bench 4 | — | 1223 |
| LiveBench Language | 59.9% | — |
Frequently asked questions
Is Claude 3.7 Sonnet better than Claude Opus 4.6?
Claude Opus 4.6 is the stronger model overall, scoring 58.3 to 39.4 on the Noometry Index.
Is Claude 3.7 Sonnet or Claude Opus 4.6 better for coding?
Claude Opus 4.6 scores higher on coding benchmarks: 57.7 versus 41.1 in the Noometry coding category.
How many benchmarks do Claude 3.7 Sonnet and Claude Opus 4.6 share?
35 benchmarks have published results for both models. Claude 3.7 Sonnet has 58 scored results on Noometry and Claude Opus 4.6 has 68.