Gemini 3.7 Flash vs GPT-6 Sol
Gemini 3.7 Flash 
GPT-6 Sol Verdict
- GPT-6 Sol (max) leads on quality: 67.7 vs 64.1.
- Gemini 3.7 Flash (high) is stronger in agents & tools, coding, human preference, maths, multimodal.
- GPT-6 Sol (max) is stronger in composite, knowledge, long context, reasoning.
- Gemini 3.7 Flash (high) is 2.7× cheaper ($1.50 vs $4.00 per 1M blended).
- Gemini 3.7 Flash (high) streams 2.2× faster (283 vs 126 tokens per second).
| Metric | Gemini 3.7 Flash (high) | GPT-6 Sol (max) |
|---|---|---|
| BenchLeader Index | 64.1 | 67.7 |
| Agents & tools score | 47.7 | – |
| Coding score | 64.7 | – |
| Composite score | 69.4 | 86.2 |
| Human preference score | 69.4 | – |
| Knowledge score | 68.9 | 75.4 |
| Long context score | 66.7 | 67.7 |
| Maths score | 65.8 | – |
| Multimodal score | 69.4 | 67.1 |
| Reasoning score | 68.9 | 95.0 |
| Blended price $/M | $1.50 | $4.00 |
| Output speed | 283 tok/s | 126 tok/s |
| Time to first answer | 12.9 s | 107.2 s |
| Context window | 1.0M | 1.1M |
| GPQA Diamond | 94.8% | – |
| FrontierMath Tiers 1–3 | 71.6% | – |
| FrontierMath Tier 4 | 36.6% | – |
| OTIS Mock AIME | 97.2% | – |
| SimpleQA Verified | 69.2% | – |
| Terminal-Bench | 11.2% | – |
| SciCode | 56.8% | – |
| LMArena Text | 1490 | – |
| LMArena Hard Prompts | 1507 | – |
| LMArena Coding | 1521 | – |
| LMArena WebDev | 1587 | – |
| LMArena Agent | -0.6 | – |
| LiveBench | 78.8% | – |
| LiveBench Reasoning | 87.8% | – |
| LiveBench Coding | 78.9% | – |
| LiveBench Agentic Coding | 58.3% | – |
| LiveBench Mathematics | 93.5% | – |
| LiveBench Data Analysis | 68.0% | – |
| LiveBench Language | 85.5% | – |
| LiveBench Instruction Following | 79.9% | – |
| AA Intelligence Index | 39.1 | 47.5 |
| AA-LCR | 81.7% | 83.7% |
| MMMU-Pro | 85.5% | 83.3% |
| AA-Omniscience | 26.5 | 27.1 |
| GPQA Diamond (AA) | 94.5% | – |
| Humanity's Last Exam (AA) | 47.9% | 47.9% |
| SciCode (AA) | 57.2% | 57.6% |
| LiveCodeBench | 88.7% | – |
| MMLU-Pro | 90.1% | – |
| IOI | 67.8% | – |
| LegalBench | 87.3% | – |
| TaxEval | 74.7% | – |
| Terminal-Bench 2.1 (Vals) | 77.5% | – |
| SWE-bench (Vals) | 80.8% | – |
| GPQA Diamond (Vals) | 93.9% | – |
| Vals Index | 59.3 | – |
| ARC-AGI-1 | 95.5% | – |
| ARC-AGI-2 | 84.6% | – |
| CritPt | 14.3% | 30.9% |
| GDPval (AA) | 43.6% | 49.4% |
| τ²-Bench Banking (AA) | 32.8% | – |
| Analyst Agent (AA) | 60.0% | – |
| Code Migration | 34.8% | – |
| Excel Modeling Benchmark | 71.3% | – |
| Finance Agent v2 | 59.0% | – |
| Harvey's Legal Agent Benchmark | 8.8% | – |
| Legal Research Bench | 34.6% | – |
| MedCode | 53.4% | – |
| MedScribe | 83.9% | – |
| MMMU-Pro (Vals) | 89.0% | – |
| MortgageTax | 66.7% | – |
| ProgramBench | 0.0% | – |
| SAGE | 49.2% | – |
| SkillsBench | 65.9% | – |
| SREBench | 4.6% | – |
| Tax Agent Bench | 57.7% | – |
| Terminal-Bench 4.0 (Vals) | 6.1% | – |
| Terminal-Bench Science | 5.7% | – |
| Vibe Code Bench v1.1 | 70.4% | – |
| LMArena Maths | 1522 | – |
| LMArena Creative Writing | 1495 | – |
| LMArena Instruction Following | 1484 | – |
| LMArena Multi-turn | 1498 | – |
| LMArena Longer Queries | 1498 | – |
| Chess Puzzles | 47.0% | – |
| Mystery Game Puzzles | 37.0% | – |
| DeepSWE | 65.3% | – |
| LMCA | 50.4% | – |
| DTBench | 96.8% | – |
| ALE-Bench | 904.3 | – |
| GDP.pdf | 23.8% | – |
| FrontierSWE | 20.3% | – |
| Terminal-Bench 4.0 (AA) | 13.6% | 43.9% |
| Terminal-Bench 2.1 (AA) | 85.8% | – |
| AutomationBench | – | 61.6% |
| GDP.pdf | – | 24.8% |
| MLCR | – | 16.1% |
| AA-Omniscience: accuracy | 55.3% | 54.5% |
| AA-Omniscience: non-hallucination | 35.5% | 39.9% |
| AA-Briefcase | – | 1483 |
Data as of 2026-09-23. Best configuration of each model; every score links to its source on the model pages.
Gemini 3.7 Flash vs GPT-6 Sol: questions
- Is Gemini 3.7 Flash better than GPT-6 Sol?
- GPT-6 Sol (max) leads on quality: 67.7 vs 64.1. The BenchLeader Index combines every independent quality benchmark; GPT-6 Sol (max) is ahead overall as of 2026-09-23, but check the category scores for your use.
- Which is cheaper, Gemini 3.7 Flash or GPT-6 Sol?
- Gemini 3.7 Flash is cheaper: $1.50 against $4.00 per million tokens, blended at three input tokens per output token.
- Which is faster, Gemini 3.7 Flash or GPT-6 Sol?
- Gemini 3.7 Flash streams faster: 283 against 126 output tokens per second.
- Which has the larger context window?
- GPT-6 Sol accepts more context: 1.1M against 1.0M tokens.