Model comparison
Claude Sonnet 5 vs GPT-6 Luna
Claude Sonnet 5 is the stronger model overall, scoring 54.6 to 53.3 on the Noometry Index. GPT-6 Luna costs 20× less per token, which makes it the better buy when Claude Sonnet 5's lead doesn't matter for your workload.
Last verified . 38 shared benchmarks.
Summary
- They share 38 benchmarks with published results for both. Claude Sonnet 5 scores higher in 6 categories and GPT-6 Luna in 4 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Claude Sonnet 5 leads 69.2 to 58.3.
- The biggest single-benchmark swing is Mystery Game Puzzles: 35% for Claude Sonnet 5 and 7% for GPT-6 Luna.
- GPT-6 Luna is cheaper at $0.10 / $0.50 per million input/output tokens, against $2 / $10 for Claude Sonnet 5.
- GPT-6 Luna accepts more context: 1.05M tokens versus 1M.
Side by side
| Claude Sonnet 5 | GPT-6 Luna | |
|---|---|---|
| Provider | Anthropic | OpenAI |
| Noometry Index | 54.6 | 53.3 |
| Released | 2026-06-29 | 2026-09-22 |
| Weights | Proprietary | Proprietary |
| Context window | 1M | 1.05M |
| Max output | 128K | 128K |
| Input $ / M tokens | $2 | $0.10 |
| Output $ / M tokens | $10 | $0.50 |
| Results tracked | 51 | 42 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
Claude Sonnet 5: 55.5 (#26), GPT-6 Luna: 55.5 (#25)
| Benchmark | Claude Sonnet 5 | GPT-6 Luna |
|---|---|---|
| DeepSWE | 53.8% | 66.6% |
| FrontierCode | 42.7% | 42.4% |
| LMArena WebDev | 1541 | 1581 |
| SciCode | 54.3% | 54.6% |
| LMArena Coding | 1483 | 1439 |
| ALE-Bench | 1,463 | 1,577 |
| CursorBench | 34.1% | — |
| GSO | 37.3% | — |
| WeirdML | 68.8% | — |
Agentic & Tool Use Claude Sonnet 5 leads
Claude Sonnet 5: 42.8 (#18), GPT-6 Luna: 33.3 (#54)
| Benchmark | Claude Sonnet 5 | GPT-6 Luna |
|---|---|---|
| APEX-Agents | 54.5% | 44.3% |
| GBAEval | 65.3% | — |
| GDP.pdf | — | 23% |
| LMArena Search | 1194 | — |
| Vending-Bench 2 | 6,378 | — |
Reasoning Too close to call
Claude Sonnet 5: 49.1 (#39), GPT-6 Luna: 48.2 (#41)
| Benchmark | Claude Sonnet 5 | GPT-6 Luna |
|---|---|---|
| NYT Connections (extended) | 75.1% | 68.7% |
| CritPt | 16.9% | 19.4% |
| Chess Puzzles | 35% | 31% |
| LMArena Hard Prompts | 1461 | 1411 |
| Mystery Game Puzzles | 35% | 7% |
| DTBench | 92.5% | 90.1% |
| LMCA | 50% | 44.5% |
| Epoch Capabilities Index | 156.21 | 156.28 |
| ARC-AGI-2 | — | 59.3% |
| SimpleBench | 60.6% | — |
| ARC-AGI-1 | — | 86.7% |
| Surface Evolver Bench | 60% | — |
| Bench to the Future 3 | 0.14 | — |
| ForecastBench | 61.1 | — |
Math GPT-6 Luna leads
Claude Sonnet 5: 66.2 (#27), GPT-6 Luna: 76.1 (#15)
| Benchmark | Claude Sonnet 5 | GPT-6 Luna |
|---|---|---|
| FrontierMath (Tiers 1-3) | 65.6% | 78.9% |
| FrontierMath Tier 4 | 29.3% | 56.1% |
| OTIS Mock AIME 2024-2025 | 94.7% | 98.9% |
| ProofBench | 77% | 64% |
| LMArena Math | 1467 | 1416 |
Knowledge GPT-6 Luna leads
Claude Sonnet 5: 55.6 (#47), GPT-6 Luna: 57.0 (#41)
| Benchmark | Claude Sonnet 5 | GPT-6 Luna |
|---|---|---|
| GPQA Diamond | 90.5% | 90.5% |
| SimpleQA Verified | 33.7% | 41.4% |
| LMArena Expert | 1490 | 1444 |
Multimodal Too close to call
Claude Sonnet 5: 42.4 (#31), GPT-6 Luna: 42.4 (#30)
| Benchmark | Claude Sonnet 5 | GPT-6 Luna |
|---|---|---|
| LMArena Vision | 1274 | 1217 |
| Blueprint-Bench 2 | 24.9% | 31.2% |
| Furniture Assembly | — | 44.2% |
| LMArena Document | 1466 | — |
Multilingual Claude Sonnet 5 leads
Claude Sonnet 5: 53.8 (#55), GPT-6 Luna: 50.5 (#117)
| Benchmark | Claude Sonnet 5 | GPT-6 Luna |
|---|---|---|
| LMArena Non-English | 1431 | 1386 |
| LMArena Chinese | 1477 | 1433 |
| LMArena French | 1460 | 1420 |
| LMArena German | 1440 | 1369 |
| LMArena Japanese | 1422 | 1369 |
| LMArena Korean | 1411 | 1360 |
| LMArena Russian | 1451 | 1394 |
| LMArena Spanish | 1437 | 1393 |
Instruction Following Claude Sonnet 5 leads
Claude Sonnet 5: 76.3 (#41), GPT-6 Luna: 74.3 (#99)
| Benchmark | Claude Sonnet 5 | GPT-6 Luna |
|---|---|---|
| LMArena Instruction Following | 1452 | 1409 |
Long Context Claude Sonnet 5 leads
Claude Sonnet 5: 44.8 (#55), GPT-6 Luna: 43.0 (#111)
| Benchmark | Claude Sonnet 5 | GPT-6 Luna |
|---|---|---|
| LMArena Longer Query | 1463 | 1409 |
Writing & Preference Claude Sonnet 5 leads
Claude Sonnet 5: 69.2 (#25), GPT-6 Luna: 58.3 (#119)
| Benchmark | Claude Sonnet 5 | GPT-6 Luna |
|---|---|---|
| LMArena Text | 1442 | 1391 |
| LMArena Creative Writing | 1416 | 1363 |
| LMArena Multi-Turn | 1454 | 1396 |
| EQ-Bench Creative Writing | 1794 | — |
| EQ-Bench 4 | 1236 | — |
Frequently asked questions
Is Claude Sonnet 5 better than GPT-6 Luna?
Claude Sonnet 5 is the stronger model overall, scoring 54.6 to 53.3 on the Noometry Index. GPT-6 Luna costs 20× less per token, which makes it the better buy when Claude Sonnet 5's lead doesn't matter for your workload.
Which is cheaper, Claude Sonnet 5 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 5 lists at $2 and $10.
Is Claude Sonnet 5 or GPT-6 Luna better for coding?
They score almost the same on coding (55.5 vs 55.5); test both on your own repository before choosing.
Which has the bigger context window?
GPT-6 Luna does, with 1.05M tokens against 1M.
How many benchmarks do Claude Sonnet 5 and GPT-6 Luna share?
38 benchmarks have published results for both models. Claude Sonnet 5 has 51 scored results on Noometry and GPT-6 Luna has 42.