Comparing 2 models
Gemini 3.1 Pro Preview vs GPT-6 Astra
Benchmarks through
Verdict
Gemini 3.1 Pro Preview places higher on 6 of the 9 tasks where both are ranked. Gemini 3.1 Pro Preview costs 5.0x less per input token. GPT-6 Astra accepts 1.0x more context.
- Choose Gemini 3.1 Pro Preview for Chat & Roleplay (#3), Roleplay & Character (#3), Creative Writing (#3).
- Choose GPT-6 Astra for Translation (#4), Literary Translation (#4), Everyday Translation (#4).
Built from the published task rankings below; ranks come from each task's own weighted benchmarks, so a win on one task says nothing about another.
Measured head to head
Benchmarks both models were run on
| Evaluation | Gemini 3.1 Pro Preview | GPT-6 Astra | Source |
|---|---|---|---|
| FACTS Grounding | 56.6% | 80.1% | source · 2026-10-24 |
| FrontierMath Tiers 1-3 (Epoch AI) | 59.6% ± 2.9 | 93.7% ± 1.4 | source · 2026-06-11 |
| GPQA Diamond (Epoch AI) | 94.4% ± 1.6 | 95.8% ± 1.4 | source · 2026-08-06 |
| GPQA Diamond (Epoch AI) | 94.1% ± 1.7 | 95.8% ± 1.4 | source · 2026-02-20 |
| LiveBench Agentic Coding | 44.1% | 57.3% | source · 2026-10-18 |
| LiveBench Coding | 76.5% | 80.4% | source · 2026-10-18 |
| LiveBench Data Analysis | 78.5% | 83.0% | source · 2026-10-18 |
| LiveBench Instruction Following | 79.1% | 75.6% | source · 2026-10-18 |
| LiveBench Language | 85.4% | 89.4% | source · 2026-10-18 |
| LiveBench Math | 91.0% | 96.8% | source · 2026-10-18 |
| LiveBench Reasoning | 84% | 92.7% | source · 2026-10-18 |
| LMArena Agent | -4.5 | 14.5 | source · 2026-11-11 |
| LMArena Agent Bash Recovery | -23.3 | 5.5 | source · 2026-10-08 |
| LMArena Agent Steerability | -8.2 | 14.8 | source · 2026-11-11 |
| LMArena Agent Task Outcome | -11.3 | 14 | source · 2026-11-11 |
| LMArena Agent Tool Hallucination | -0.4 | 1 | source · 2026-11-11 |
| LMArena Creative Writing | 1483 Elo | 1421 Elo | source · 2026-11-11 |
| LMArena Instruction Following | 1465 Elo | 1456 Elo | source · 2026-11-11 |
| LMArena Long Query | 1482 Elo | 1459 Elo | source · 2026-11-11 |
| LMArena Math | 1485 Elo | 1459 Elo | source · 2026-11-11 |
| LMArena Text | 1487 Elo | 1474 Elo | source · 2026-11-11 |
| LMArena Text Factuality | 1472 Elo | 1466 Elo | source · 2026-11-11 |
| LMArena Vision | 1280 Elo | 1283 Elo | source · 2026-11-09 |
| LMArena Vision Diagrams | 1309 Elo | 1311 Elo | source · 2026-11-09 |
| LMArena Vision OCR | 1295 Elo | 1297 Elo | source · 2026-11-09 |
| LMArena WebDev | 1446 Elo | 1764 Elo | source · 2026-11-11 |
| LMArena WebDev Image-to-Web | 1480 Elo | 1722 Elo | source · 2026-11-10 |
| OTIS Mock AIME (Epoch AI) | 95.6% ± 3.1 | 100% ± 0 | source · 2026-08-06 |
| OTIS Mock AIME (Epoch AI) | 95.6% ± 3.1 | 100% ± 0 | source · 2026-02-20 |
| SimpleQA Verified (Epoch AI) | 73.5% ± 1.4 | 75.6% ± 1.4 | source · 2026-08-10 |
Values are as published by each evaluator, in its own units; rows from different evaluators are not on a common scale.
task evidence
How they place when the question changes
Lower ranks are better. “Best” marks the strongest option in this shortlist for that task.