Model comparison
Claude 3.7 Sonnet vs GPT-5.6 Luna
GPT-5.6 Luna is the stronger model overall, scoring 54.6 to 39.5 on the Noometry Index.
Last verified . 25 shared benchmarks.
Summary
- They share 25 benchmarks with published results for both. Claude 3.7 Sonnet scores higher in 1 category and GPT-5.6 Luna in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.6 Luna leads 77.7 to 37.5.
- The biggest single-benchmark swing is ARC-AGI-1: 28.6% for Claude 3.7 Sonnet and 88% for GPT-5.6 Luna.
Side by side
| Claude 3.7 Sonnet | GPT-5.6 Luna | |
|---|---|---|
| Provider | Anthropic | OpenAI |
| Noometry Index | 39.5 | 54.6 |
| Released | 2025-02-24 | 2026-07-09 |
| Weights | Proprietary | Proprietary |
| Context window | — | 1.05M |
| Max output | — | 128K |
| Input $ / M tokens | — | $0.20 |
| Output $ / M tokens | — | $1.20 |
| Results tracked | 58 | 52 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.6 Luna leads
Claude 3.7 Sonnet: 40.6 (#136), GPT-5.6 Luna: 54.5 (#28)
| Benchmark | Claude 3.7 Sonnet | GPT-5.6 Luna |
|---|---|---|
| LMArena Coding | 1361 | 1466 |
| SWE-bench Verified | 61% | — |
| DeepSWE | — | 67.2% |
| FrontierCode | — | 39.8% |
| SWE-bench Verified (bash only) | 52.8% | — |
| Aider Polyglot | 64.9% | — |
| CursorBench | — | 35.9% |
| LMArena WebDev | — | 1519 |
| SciCode | — | 53.6% |
| GSO | 3.8% | — |
| WeirdML | — | 60.9% |
| LiveBench Coding | 74.5% | — |
| CadEval | 54% | — |
| ALE-Bench | — | 1,667 |
Agentic & Tool Use Too close to call
Claude 3.7 Sonnet: 34.1 (#50), GPT-5.6 Luna: 34.4 (#45)
| Benchmark | Claude 3.7 Sonnet | GPT-5.6 Luna |
|---|---|---|
| APEX-Agents | — | 43% |
| TheAgentCompany | 30.9% | — |
| Cybench | 20% | — |
| DeepResearch Bench | 43.6% | — |
| OSWorld | 35.8% | — |
| BALROG | — | 45.6% |
| GDP.pdf | — | 22.7% |
| METR Time Horizons | 60% | — |
| Vending-Bench 2 | — | 4,095 |
Reasoning GPT-5.6 Luna leads
Claude 3.7 Sonnet: 18.6 (#277), GPT-5.6 Luna: 47.6 (#43)
| Benchmark | Claude 3.7 Sonnet | GPT-5.6 Luna |
|---|---|---|
| ARC-AGI-2 | 0.9% | 59.5% |
| SimpleBench | 46.4% | 46.8% |
| ARC-AGI-1 | 28.6% | 88% |
| LMArena Hard Prompts | 1333 | 1451 |
| Epoch Capabilities Index | 141.16 | 156.39 |
| Kagi LLM Benchmark | — | 49.1% |
| NYT Connections (extended) | — | 69.4% |
| CritPt | — | 20.6% |
| Chess Puzzles | — | 40% |
| EnigmaEval | 4.2% | — |
| LiveBench Reasoning | 87.8% | — |
| Mystery Game Puzzles | — | 21% |
| DTBench | — | 89.1% |
| LiveBench Data Analysis | 74% | — |
| LMCA | — | 48.5% |
| Surface Evolver Bench | — | 61.9% |
| ForecastBench | 61.8 | — |
| LiveBench | 76.1% | — |
Math GPT-5.6 Luna leads
Claude 3.7 Sonnet: 37.5 (#153), GPT-5.6 Luna: 77.7 (#14)
| Benchmark | Claude 3.7 Sonnet | GPT-5.6 Luna |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 57.8% | 98.3% |
| LMArena Math | 1337 | 1458 |
| FrontierMath (Tiers 1-3) | — | 82.1% |
| FrontierMath Tier 4 | — | 61% |
| ProofBench | — | 60% |
| Omni-MATH | 33% | — |
| LiveBench Math | 79% | — |
| MATH Level 5 | 91.2% | — |
| FrontierMath (Feb 2025 set) | 4.1% | — |
Knowledge GPT-5.6 Luna leads
Claude 3.7 Sonnet: 39.8 (#130), GPT-5.6 Luna: 58.5 (#34)
| Benchmark | Claude 3.7 Sonnet | GPT-5.6 Luna |
|---|---|---|
| GPQA Diamond | 79.7% | 91.6% |
| LMArena Expert | 1321 | 1478 |
| Humanity's Last Exam | 8% | — |
| SimpleQA Verified | — | 41% |
| MMLU-Pro | 78.4% | — |
| Confabulations | 14.7% | — |
| GPQA (HELM) | 60.8% | — |
Multimodal GPT-5.6 Luna leads
Claude 3.7 Sonnet: 33.7 (#95), GPT-5.6 Luna: 42.7 (#28)
| Benchmark | Claude 3.7 Sonnet | GPT-5.6 Luna |
|---|---|---|
| LMArena Vision | 1169 | 1258 |
| GeoBench | 68% | — |
| VPCT | 39% | — |
| Blueprint-Bench 2 | — | 22.6% |
| Furniture Assembly | — | 42.5% |
| LMArena Document | — | 1457 |
| SpatialViz-Bench | 33.9% | — |
Multilingual GPT-5.6 Luna leads
Claude 3.7 Sonnet: 44.1 (#179), GPT-5.6 Luna: 52.8 (#78)
| Benchmark | Claude 3.7 Sonnet | GPT-5.6 Luna |
|---|---|---|
| LMArena Non-English | 1296 | 1417 |
| LMArena Chinese | 1299 | 1470 |
| LMArena French | 1303 | 1456 |
| LMArena German | 1301 | 1454 |
| LMArena Japanese | 1267 | 1411 |
| LMArena Korean | 1249 | 1415 |
| LMArena Russian | 1311 | 1428 |
| LMArena Spanish | 1298 | 1448 |
Instruction Following GPT-5.6 Luna leads
Claude 3.7 Sonnet: 72.9 (#125), GPT-5.6 Luna: 75.6 (#57)
| Benchmark | Claude 3.7 Sonnet | GPT-5.6 Luna |
|---|---|---|
| LMArena Instruction Following | 1352 | 1437 |
| LiveBench Instruction Following | 81.3% | — |
| IFEval | 83.4% | — |
Long Context Claude 3.7 Sonnet leads
Claude 3.7 Sonnet: 50.3 (#10), GPT-5.6 Luna: 43.9 (#82)
| Benchmark | Claude 3.7 Sonnet | GPT-5.6 Luna |
|---|---|---|
| LMArena Longer Query | 1373 | 1436 |
| Fiction.LiveBench | 83.3% | — |
Writing & Preference GPT-5.6 Luna leads
Claude 3.7 Sonnet: 54.4 (#150), GPT-5.6 Luna: 68.0 (#29)
| Benchmark | Claude 3.7 Sonnet | GPT-5.6 Luna |
|---|---|---|
| LMArena Text | 1314 | 1431 |
| LMArena Creative Writing | 1332 | 1396 |
| EQ-Bench Creative Writing | 1412 | 1829 |
| LMArena Multi-Turn | 1339 | 1434 |
| Short-Story Creative Writing | 81.1% | — |
| WildBench | 81.4% | — |
| EQ-Bench 4 | — | 1156 |
| LiveBench Language | 59.9% | — |
Frequently asked questions
Is Claude 3.7 Sonnet better than GPT-5.6 Luna?
GPT-5.6 Luna is the stronger model overall, scoring 54.6 to 39.5 on the Noometry Index.
Is Claude 3.7 Sonnet or GPT-5.6 Luna better for coding?
GPT-5.6 Luna scores higher on coding benchmarks: 54.5 versus 40.6 in the Noometry coding category.
How many benchmarks do Claude 3.7 Sonnet and GPT-5.6 Luna share?
25 benchmarks have published results for both models. Claude 3.7 Sonnet has 58 scored results on Noometry and GPT-5.6 Luna has 52.