Compare / Claude Sonnet 4.6 vs GPT-6 Luna
Claude Sonnet 4.6 vs GPT-6 Luna
15 results from 5 sources that measured both models. Each row is in the source's own units; there is no overall winner.
| Benchmark | Metric | Claude Sonnet 4.6 | GPT-6 Luna |
|---|---|---|---|
| APEX-Agentsreported | Tasks passed% of tasks | 43.0% | 44.3% |
| Arena (formerly LMArena)reported | OverallArena rating | 1,461 | 1,422 |
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,482 | 1,452 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,449 | 1,403 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,506 | 1,505 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,475 | 1,443 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,472 | 1,442 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,455 | 1,415 |
| Artificial Analysisreported | Artificial Analysis Intelligence Indexindex score | 30.1 | 38.1 |
| Artificial Analysisreported | Humanity's Last Exam% of questions | 33.6% | 38.5% |
| Artificial Analysisreported | Long-context reasoning (AA-LCR)% of questions | 80.0% | 83.3% |
| Artificial Analysisreported | SciCode% of problems | 50.1% | 54.6% |
| Artificial Analysisreported | Terminal-Bench 4.0% of tasks | 3.0% | 12.6% |
| GDP.pdfreported | Professional document tasks% of rubric | 18.0% | 23.0% |
| SimpleQA Verified (Epoch AI)reported | Correct answers% of questions | 35.5% | 41.4% |
Bold green marks the better value on that metric. Where a source reports ranges that overlap, the difference may not be meaningful; see the benchmark page for ranges.