Compare / Claude Opus 5.5 vs GPT-5.6 Luna
Claude Opus 5.5 vs GPT-5.6 Luna
20 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 Opus 5.5 | GPT-5.6 Luna |
|---|---|---|---|
| Arena (formerly LMArena)reported | Confirmed task successIPS effect estimate | 0.16 | -0.05 |
| Arena (formerly LMArena)reported | Praise over complaintIPS effect estimate | 0.19 | -0.04 |
| Arena (formerly LMArena)reported | SteerabilityIPS effect estimate | 0.14 | 0.02 |
| Arena (formerly LMArena)reported | Tool groundingIPS effect estimate | 0.00 | 0.00 |
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,499 | 1,463 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,521 | 1,412 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,556 | 1,491 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,516 | 1,446 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,509 | 1,454 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,520 | 1,435 |
| Artificial Analysisreported | Artificial Analysis Intelligence Indexindex score | 57.6 | 37.3 |
| Artificial Analysisreported | Humanity's Last Exam% of questions | 61.4% | 39.5% |
| Artificial Analysisreported | Long-context reasoning (AA-LCR)% of questions | 84.7% | 83.7% |
| Artificial Analysisreported | SciCode% of problems | 66.9% | 53.6% |
| Artificial Analysisreported | Terminal-Bench 4.0% of tasks | 59.6% | 11.6% |
| GDP.pdfreported | Professional document tasks% of rubric | 30.6% | 22.7% |
| SimpleQA Verified (Epoch AI)reported | Correct answers% of questions | 72.2% | 41.0% |
| UGI Leaderboardreported | Requested-length error% off the requested word count | 2.0% | 15.0% |
| UGI Leaderboardreported | Style adherencescore from 0 to 1 | 0.43 | 0.38 |
| UGI Leaderboardreported | Writing scorescore out of 100 | 70.0 | 53.7 |
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.