Gemini 3.1 Pro Preview vs Gemini 3 Pro
Côte à côte. Chaque métrique. Chaque benchmark.
| Type | Gemini 3.1 Pro Preview | Gemini 3 Pro |
|---|---|---|
| Provider |  Google DeepMind | Google DeepMind |
| score moyen | 60.6 | 60.5 |
| Prix d'entrée | $2.00 | - |
| Prix de sortie | $12.00 | - |
| Fenêtre de contexte | 1.0M tokens (~524 books) | - |
| Sorti le | 2026-02-19 | 2024-01-01 |
| Code source ouvert | Proprietary | Proprietary |
Scores de benchmark
19 benchmarks · Gemini 3.1 Pro Preview: 17, Gemini 3 Pro: 2
| Benchmark | Catégorie | Gemini 3.1 Pro Preview | Gemini 3 Pro |
|---|---|---|---|
| APEX-Agents | agentic | 33.5 | 18.4 |
| ARC-AGI | reasoning | 98.0 | 75.0 |
| ARC-AGI-2 | reasoning | 77.1 | 31.1 |
| Artificial Analysis — Agentic Index | speed | 59.1 | 45.0 |
| Artificial Analysis — Coding Index | speed | 55.5 | 39.4 |
| Artificial Analysis — Quality Index | speed | 57.2 | 41.3 |
| Chatbot Arena Elo — Coding | arena | 1455.7 | 1437.6 |
| Chatbot Arena Elo — Overall | arena | 1492.6 | 1486.2 |
| Chess Puzzles | knowledge | 55.0 | 31.0 |
| FrontierMath-2025-02-28-Private | math | 36.9 | 37.6 |
| FrontierMath-Tier-4-2025-07-01-Private | math | 16.7 | 18.8 |
| GPQA diamond | knowledge | 92.1 | 90.2 |
| OTIS Mock AIME 2024-2025 | math | 95.6 | 91.4 |
| PostTrainBench | knowledge | 21.6 | 18.1 |
| SimpleBench | reasoning | 75.5 | 71.7 |
| SimpleQA Verified | knowledge | 77.3 | 72.9 |
| SWE-Bench verified | coding | 75.6 | 72.9 |
| Terminal Bench | coding | 78.4 | 69.4 |
| WeirdML | coding | 72.1 | 69.9 |