Model comparison
Claude Sonnet 4 vs GPT-6 Luna
GPT-6 Luna is the stronger model overall, scoring 45.1 to 40.6 on the Noometry Index.
Last verified . 13 shared benchmarks.
Summary
- They share 13 benchmarks with published results for both. Claude Sonnet 4 scores higher in 2 categories and GPT-6 Luna in 7 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in multimodal, where GPT-6 Luna leads 42.0 to 25.6.
- GPT-6 Luna is cheaper at $0.10 / $0.50 per million input/output tokens, against $3 / $15 for Claude Sonnet 4.
- GPT-6 Luna accepts more context: 1.05M tokens versus 200K.
Side by side
| Claude Sonnet 4 | GPT-6 Luna | |
|---|---|---|
| Provider | Anthropic | OpenAI |
| Noometry Index | 40.6 | 45.1 |
| Released | 2025-05-22 | 2026-09-22 |
| Weights | Proprietary | Proprietary |
| Context window | 200K | 1.05M |
| Max output | 64K | 128K |
| Input $ / M tokens | $3 | $0.10 |
| Output $ / M tokens | $15 | $0.50 |
| Results tracked | 58 | 18 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-6 Luna leads
Claude Sonnet 4: 44.0 (#70), GPT-6 Luna: 54.7 (#24)
| Benchmark | Claude Sonnet 4 | GPT-6 Luna |
|---|---|---|
| LMArena Coding | 1414 | 1448 |
| DeepSWE | — | 66.6% |
| FrontierCode | — | 42.4% |
| SWE-bench Verified (bash only) | 64.9% | — |
| Aider Polyglot | 61.3% | — |
| LMArena WebDev | — | 1583 |
| SciCode | 40% | — |
| GSO | 4.9% | — |
| WeirdML | 46.1% | — |
| ALE-Bench | 655.35 | — |
Agentic & Tool Use Not comparable
Claude Sonnet 4: 38.3 (#29), GPT-6 Luna: —
| Benchmark | Claude Sonnet 4 | GPT-6 Luna |
|---|---|---|
| TheAgentCompany | 33.1% | — |
| Cybench | 35% | — |
| DeepResearch Bench | 46.6% | — |
| OSWorld | 43.9% | — |
| METR Time Horizons | 62% | — |
Reasoning GPT-6 Luna leads
Claude Sonnet 4: 23.0 (#169), GPT-6 Luna: 31.7 (#68)
| Benchmark | Claude Sonnet 4 | GPT-6 Luna |
|---|---|---|
| LMArena Hard Prompts | 1372 | 1415 |
| ARC-AGI-2 | 5.9% | — |
| SimpleBench | 45.5% | — |
| Kagi LLM Benchmark | 73% | — |
| NYT Connections (extended) | — | 68.7% |
| ARC-AGI-1 | 40% | — |
| CritPt | 0.3% | — |
| EnigmaEval | 3.1% | — |
| DTBench | 77.1% | — |
| LMCA | 29% | — |
| Epoch Capabilities Index | 141.69 | — |
| ForecastBench | 60.2 | — |
Math Claude Sonnet 4 leads
Claude Sonnet 4: 41.5 (#71), GPT-6 Luna: 35.6 (#140)
| Benchmark | Claude Sonnet 4 | GPT-6 Luna |
|---|---|---|
| LMArena Math | 1375 | 1396 |
| OTIS Mock AIME 2024-2025 | 71.1% | — |
| Omni-MATH | 60.2% | — |
| MATH Level 5 | 84.4% | — |
| FrontierMath (Feb 2025 set) | 4.1% | — |
| FrontierMath Tier 4 (v1) | 0% | — |
Knowledge Claude Sonnet 4 leads
Claude Sonnet 4: 41.6 (#99), GPT-6 Luna: 40.1 (#110)
| Benchmark | Claude Sonnet 4 | GPT-6 Luna |
|---|---|---|
| LMArena Expert | 1371 | 1458 |
| GPQA Diamond | 79.2% | — |
| Humanity's Last Exam | 7.8% | — |
| MMLU-Pro | 84.3% | — |
| Confabulations | 13.2% | — |
| Vectara Hallucination Rate | 10.3% | — |
| GPQA (HELM) | 70.6% | — |
Multimodal GPT-6 Luna leads
Claude Sonnet 4: 25.6 (#116), GPT-6 Luna: 42.0 (#27)
| Benchmark | Claude Sonnet 4 | GPT-6 Luna |
|---|---|---|
| LMArena Vision | 1191 | — |
| GeoBench | 37% | — |
| VPCT | 34% | — |
| Furniture Assembly | — | 44.2% |
| MindCube | 44.8% | — |
Multilingual GPT-6 Luna leads
Claude Sonnet 4: 46.9 (#151), GPT-6 Luna: 51.0 (#111)
| Benchmark | Claude Sonnet 4 | GPT-6 Luna |
|---|---|---|
| LMArena Non-English | 1333 | 1390 |
| LMArena Chinese | 1351 | 1425 |
| LMArena Russian | 1355 | 1392 |
| LMArena Spanish | 1356 | 1416 |
| LMArena French | 1364 | — |
| LMArena German | 1331 | — |
| LMArena Japanese | 1301 | — |
| LMArena Korean | 1293 | — |
Instruction Following GPT-6 Luna leads
Claude Sonnet 4: 71.8 (#141), GPT-6 Luna: 74.7 (#85)
| Benchmark | Claude Sonnet 4 | GPT-6 Luna |
|---|---|---|
| LMArena Instruction Following | 1375 | 1414 |
| IFEval | 84% | — |
Long Context GPT-6 Luna leads
Claude Sonnet 4: 33.6 (#254), GPT-6 Luna: 43.3 (#95)
| Benchmark | Claude Sonnet 4 | GPT-6 Luna |
|---|---|---|
| LMArena Longer Query | 1398 | 1417 |
| Fiction.LiveBench | 46.9% | — |
Writing & Preference GPT-6 Luna leads
Claude Sonnet 4: 57.3 (#127), GPT-6 Luna: 58.4 (#115)
| Benchmark | Claude Sonnet 4 | GPT-6 Luna |
|---|---|---|
| LMArena Text | 1351 | 1392 |
| LMArena Creative Writing | 1345 | 1357 |
| LMArena Multi-Turn | 1376 | 1400 |
| Short-Story Creative Writing | 81.4% | — |
| EQ-Bench Creative Writing | 1483 | — |
| WildBench | 83.8% | — |
Frequently asked questions
Is Claude Sonnet 4 better than GPT-6 Luna?
GPT-6 Luna is the stronger model overall, scoring 45.1 to 40.6 on the Noometry Index.
Which is cheaper, Claude Sonnet 4 or GPT-6 Luna?
GPT-6 Luna is cheaper. It lists at $0.10 per million input tokens and $0.50 per million output tokens; Claude Sonnet 4 lists at $3 and $15.
Is Claude Sonnet 4 or GPT-6 Luna better for coding?
GPT-6 Luna scores higher on coding benchmarks: 54.7 versus 44.0 in the Noometry coding category.
Which has the bigger context window?
GPT-6 Luna does, with 1.05M tokens against 200K.
How many benchmarks do Claude Sonnet 4 and GPT-6 Luna share?
13 benchmarks have published results for both models. Claude Sonnet 4 has 58 scored results on Noometry and GPT-6 Luna has 18.