Model comparison
Claude Sonnet 4.6 vs GPT-5.6 Luna
GPT-5.6 Luna is the stronger model overall, scoring 54.6 to 50.3 on the Noometry Index.
Last verified . 44 shared benchmarks.
Summary
- They share 44 benchmarks with published results for both. Claude Sonnet 4.6 scores higher in 5 categories and GPT-5.6 Luna in 5 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.6 Luna leads 77.7 to 52.9.
- The biggest single-benchmark swing is DeepSWE: 29.9% for Claude Sonnet 4.6 and 67.2% for GPT-5.6 Luna.
- GPT-5.6 Luna is cheaper at $0.20 / $1.20 per million input/output tokens, against $3 / $15 for Claude Sonnet 4.6.
- GPT-5.6 Luna accepts more context: 1.05M tokens versus 1M.
Side by side
| Claude Sonnet 4.6 | GPT-5.6 Luna | |
|---|---|---|
| Provider | Anthropic | OpenAI |
| Noometry Index | 50.3 | 54.6 |
| Released | 2026-02-17 | 2026-07-09 |
| Weights | Proprietary | Proprietary |
| Context window | 1M | 1.05M |
| Max output | 128K | 128K |
| Input $ / M tokens | $3 | $0.20 |
| Output $ / M tokens | $15 | $1.20 |
| Results tracked | 57 | 52 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.6 Luna leads
Claude Sonnet 4.6: 46.3 (#67), GPT-5.6 Luna: 54.5 (#28)
| Benchmark | Claude Sonnet 4.6 | GPT-5.6 Luna |
|---|---|---|
| DeepSWE | 29.9% | 67.2% |
| FrontierCode | 24.3% | 39.8% |
| LMArena WebDev | 1522 | 1519 |
| SciCode | 46.8% | 53.6% |
| WeirdML | 66.1% | 60.9% |
| LMArena Coding | 1504 | 1466 |
| ALE-Bench | 1,327 | 1,667 |
| SWE-bench Verified | 75.2% | — |
| CursorBench | — | 35.9% |
Agentic & Tool Use Claude Sonnet 4.6 leads
Claude Sonnet 4.6: 39.1 (#28), GPT-5.6 Luna: 34.4 (#45)
| Benchmark | Claude Sonnet 4.6 | GPT-5.6 Luna |
|---|---|---|
| APEX-Agents | 43% | 43% |
| GDP.pdf | 18% | 22.7% |
| Vending-Bench 2 | 7,204 | 4,095 |
| Terminal-Bench | 53.4% | — |
| OSWorld 2.0 | 9.3% | — |
| DeepResearch Bench | 54.9% | — |
| OSWorld | 72.1% | — |
| BALROG | — | 45.6% |
| ExploitBench | 23.6% | — |
| GBAEval | 48.8% | — |
| LMArena Search | 1221 | — |
Reasoning GPT-5.6 Luna leads
Claude Sonnet 4.6: 46.1 (#45), GPT-5.6 Luna: 47.6 (#43)
| Benchmark | Claude Sonnet 4.6 | GPT-5.6 Luna |
|---|---|---|
| ARC-AGI-2 | 60.4% | 59.5% |
| NYT Connections (extended) | 80.9% | 69.4% |
| ARC-AGI-1 | 86.5% | 88% |
| CritPt | 3.1% | 20.6% |
| Chess Puzzles | 13% | 40% |
| LMArena Hard Prompts | 1484 | 1451 |
| Mystery Game Puzzles | 16% | 21% |
| DTBench | 89.9% | 89.1% |
| LMCA | 46.5% | 48.5% |
| Epoch Capabilities Index | 152.24 | 156.39 |
| SimpleBench | — | 46.8% |
| Kagi LLM Benchmark | — | 49.1% |
| Thematic Generalization | 76.3% | — |
| Surface Evolver Bench | — | 61.9% |
| ForecastBench | 62 | — |
Math GPT-5.6 Luna leads
Claude Sonnet 4.6: 52.9 (#49), GPT-5.6 Luna: 77.7 (#14)
| Benchmark | Claude Sonnet 4.6 | GPT-5.6 Luna |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 85.8% | 98.3% |
| ProofBench | 45% | 60% |
| LMArena Math | 1462 | 1458 |
| FrontierMath (Tiers 1-3) | — | 82.1% |
| FrontierMath Tier 4 | — | 61% |
| FrontierMath (Feb 2025 set) | 32.4% | — |
| FrontierMath Tier 4 (v1) | 8.3% | — |
Knowledge GPT-5.6 Luna leads
Claude Sonnet 4.6: 51.7 (#65), GPT-5.6 Luna: 58.5 (#34)
| Benchmark | Claude Sonnet 4.6 | GPT-5.6 Luna |
|---|---|---|
| GPQA Diamond | 87.4% | 91.6% |
| SimpleQA Verified | 35.5% | 41% |
| LMArena Expert | 1500 | 1478 |
| Vectara Hallucination Rate | 10.6% | — |
Multimodal GPT-5.6 Luna leads
Claude Sonnet 4.6: 38.0 (#68), GPT-5.6 Luna: 42.7 (#28)
| Benchmark | Claude Sonnet 4.6 | GPT-5.6 Luna |
|---|---|---|
| LMArena Vision | 1283 | 1258 |
| Blueprint-Bench 2 | 6.7% | 22.6% |
| LMArena Document | 1482 | 1457 |
| Furniture Assembly | — | 42.5% |
Multilingual Claude Sonnet 4.6 leads
Claude Sonnet 4.6: 54.4 (#41), GPT-5.6 Luna: 52.8 (#78)
| Benchmark | Claude Sonnet 4.6 | GPT-5.6 Luna |
|---|---|---|
| LMArena Non-English | 1440 | 1417 |
| LMArena Chinese | 1491 | 1470 |
| LMArena French | 1465 | 1456 |
| LMArena German | 1428 | 1454 |
| LMArena Japanese | 1420 | 1411 |
| LMArena Korean | 1411 | 1415 |
| LMArena Russian | 1440 | 1428 |
| LMArena Spanish | 1464 | 1448 |
Instruction Following Claude Sonnet 4.6 leads
Claude Sonnet 4.6: 77.4 (#25), GPT-5.6 Luna: 75.6 (#57)
| Benchmark | Claude Sonnet 4.6 | GPT-5.6 Luna |
|---|---|---|
| LMArena Instruction Following | 1475 | 1437 |
Long Context Claude Sonnet 4.6 leads
Claude Sonnet 4.6: 45.3 (#44), GPT-5.6 Luna: 43.9 (#82)
| Benchmark | Claude Sonnet 4.6 | GPT-5.6 Luna |
|---|---|---|
| LMArena Longer Query | 1479 | 1436 |
Writing & Preference Claude Sonnet 4.6 leads
Claude Sonnet 4.6: 70.2 (#22), GPT-5.6 Luna: 68.0 (#29)
| Benchmark | Claude Sonnet 4.6 | GPT-5.6 Luna |
|---|---|---|
| LMArena Text | 1458 | 1431 |
| LMArena Creative Writing | 1435 | 1396 |
| EQ-Bench Creative Writing | 1810 | 1829 |
| EQ-Bench 4 | 1207 | 1156 |
| LMArena Multi-Turn | 1464 | 1434 |
Frequently asked questions
Is Claude Sonnet 4.6 better than GPT-5.6 Luna?
GPT-5.6 Luna is the stronger model overall, scoring 54.6 to 50.3 on the Noometry Index.
Which is cheaper, Claude Sonnet 4.6 or GPT-5.6 Luna?
GPT-5.6 Luna is cheaper. It lists at $0.20 per million input tokens and $1.20 per million output tokens; Claude Sonnet 4.6 lists at $3 and $15.
Is Claude Sonnet 4.6 or GPT-5.6 Luna better for coding?
GPT-5.6 Luna scores higher on coding benchmarks: 54.5 versus 46.3 in the Noometry coding category.
Which has the bigger context window?
GPT-5.6 Luna does, with 1.05M tokens against 1M.
How many benchmarks do Claude Sonnet 4.6 and GPT-5.6 Luna share?
44 benchmarks have published results for both models. Claude Sonnet 4.6 has 57 scored results on Noometry and GPT-5.6 Luna has 52.