Compare / Claude Sonnet 4.6 vs Gemini 3.5 Flash Lite
Claude Sonnet 4.6 vs Gemini 3.5 Flash Lite
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 | Claude Sonnet 4.6 | Gemini 3.5 Flash Lite |
|---|---|---|---|
| Arena (formerly LMArena)reported | OverallArena rating | 1,461 | 1,448 |
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,482 | 1,454 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,449 | 1,435 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,506 | 1,467 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,475 | 1,444 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,472 | 1,456 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,455 | 1,442 |
| Artificial Analysisreported | Artificial Analysis Coding Indexindex score | 63.0 | 49.3 |
| Artificial Analysisreported | Artificial Analysis Intelligence Indexindex score | 30.1 | 22.2 |
| Artificial Analysisreported | GPQA Diamond% of questions | 87.5% | 83.8% |
| Artificial Analysisreported | Humanity's Last Exam% of questions | 33.6% | 18.8% |
| Artificial Analysisreported | Long-context reasoning (AA-LCR)% of questions | 80.0% | 76.0% |
| Artificial Analysisreported | SciCode% of problems | 50.1% | 41.3% |
| Artificial Analysisreported | Terminal-Bench 2.1% of tasks | 71.2% | 53.6% |
| Artificial Analysisreported | Terminal-Bench 4.0% of tasks | 3.0% | 1.0% |
| Artificial Analysisreported | Τ-bench banking% of tasks | 34.4% | 17.5% |
| GDP.pdfreported | Professional document tasks% of rubric | 18.0% | 10.0% |
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.