Compare / Gemini 3.5 Flash Lite vs GPT-5.6 Luna
Gemini 3.5 Flash Lite vs GPT-5.6 Luna
17 results from 3 sources that measured both models. Each row is in the source's own units; there is no overall winner.
| Benchmark | Metric | Gemini 3.5 Flash Lite | GPT-5.6 Luna |
|---|---|---|---|
| Arena (formerly LMArena)reported | OverallArena rating | 1,448 | 1,461 |
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,454 | 1,463 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,435 | 1,412 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,467 | 1,491 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,444 | 1,446 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,456 | 1,454 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,442 | 1,435 |
| Artificial Analysisreported | Artificial Analysis Coding Indexindex score | 49.3 | 71.4 |
| Artificial Analysisreported | Artificial Analysis Intelligence Indexindex score | 22.2 | 37.3 |
| Artificial Analysisreported | GPQA Diamond% of questions | 83.8% | 91.1% |
| Artificial Analysisreported | Humanity's Last Exam% of questions | 18.8% | 39.5% |
| Artificial Analysisreported | Long-context reasoning (AA-LCR)% of questions | 76.0% | 83.7% |
| Artificial Analysisreported | SciCode% of problems | 41.3% | 53.6% |
| Artificial Analysisreported | Terminal-Bench 2.1% of tasks | 53.6% | 80.9% |
| Artificial Analysisreported | Terminal-Bench 4.0% of tasks | 1.0% | 11.6% |
| Artificial Analysisreported | Τ-bench banking% of tasks | 17.5% | 31.1% |
| GDP.pdfreported | Professional document tasks% of rubric | 10.0% | 22.7% |
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.