Gemini 3.5 Flash vs GPT-5.6 Sol
Gemini 3.5 Flash 
GPT-5.6 Sol Verdict
- GPT-5.6 Sol (high) leads on quality: 68.0 vs 63.9.
- Gemini 3.5 Flash (medium) is stronger in human preference, instruction following, knowledge, multimodal, reasoning.
- GPT-5.6 Sol (high) is stronger in agents & tools, coding, composite, long context.
- Gemini 3.5 Flash (medium) is 2.4× cheaper ($3.38 vs $8.00 per 1M blended).
- Gemini 3.5 Flash (medium) streams 3.5× faster (209 vs 59 tokens per second).
| Metric | Gemini 3.5 Flash (medium) | GPT-5.6 Sol (high) |
|---|---|---|
| BenchLeader Index | 63.9 | 68.0 |
| Agents & tools score | 67.8 | 87.6 |
| Coding score | 58.9 | 72.0 |
| Composite score | 71.4 | 82.5 |
| Human preference score | 67.4 | – |
| Instruction following score | 72.2 | 67.6 |
| Knowledge score | 73.9 | 73.7 |
| Long context score | 63.6 | 67.4 |
| Multimodal score | 67.6 | 66.5 |
| Reasoning score | 66.8 | 62.9 |
| Blended price $/M | $3.38 | $8.00 |
| Output speed | 209 tok/s | 59 tok/s |
| Time to first answer | 17.0 s | 31.4 s |
| Context window | 1M | 1M |
| SciCode | – | 56.9% |
| WeirdML | – | 88.8% |
| LMArena Text | 1476 | – |
| LMArena Hard Prompts | 1493 | – |
| LMArena Coding | 1503 | – |
| LMArena WebDev | 1491 | – |
| LMArena Vision | 1306 | – |
| AA Intelligence Index | 33.6 | 42.5 |
| IFBench | 74.6% | 69.2% |
| AA-LCR | 74.3% | 81.7% |
| MMMU-Pro | 83.9% | 81.8% |
| AA-Omniscience | 20.8 | 20.4 |
| Terminal-Bench Hard | 39.4% | 62.1% |
| GPQA Diamond (AA) | 92.1% | 92.8% |
| Humanity's Last Exam (AA) | 41.3% | 46.0% |
| SciCode (AA) | – | 57.8% |
| τ²-Bench Telecom (AA) | 95.6% | 83.3% |
| EnigmaEval | – | 37.1% |
| ARC-AGI-1 | – | 97.0% |
| ARC-AGI-2 | – | 85.4% |
| ARC-AGI-3 | – | 2.2% |
Data as of 2026-09-09. Best configuration of each model; every score links to its source on the model pages.