Gemma 4 31B vs GPT-5.1-Codex-Max
Côte à côte. Chaque métrique. Chaque benchmark.
| Type | Gemma 4 31B | GPT-5.1-Codex-Max |
|---|---|---|
| Provider |  Google DeepMind |  OpenAI |
| score moyen | 61.6 | 72.0 |
| Prix d'entrée | $0.13 | $1.25 |
| Prix de sortie | $0.38 | $10.00 |
| Fenêtre de contexte | 262K tokens (~131 books) | 400K tokens (~200 books) |
| Sorti le | 2026-04-02 | 2025-12-04 |
| Code source ouvert | Open Source | Proprietary |
Scores de benchmark
8 benchmarks · Gemma 4 31B: 2, GPT-5.1-Codex-Max: 6
| Benchmark | Catégorie | Gemma 4 31B | GPT-5.1-Codex-Max |
|---|---|---|---|
| LiveBench — Agentic Coding | coding | 40.0 | 56.7 |
| LiveBench — Coding | coding | 60.3 | 81.4 |
| LiveBench — Data Analysis | reasoning | 58.8 | 54.9 |
| LiveBench — If | language | 67.6 | 67.1 |
| LiveBench — Language | language | 71.3 | 75.4 |
| LiveBench — Mathematics | math | 73.9 | 83.7 |
| LiveBench — Overall | knowledge | 61.6 | 72.0 |
| LiveBench — Reasoning | reasoning | 59.4 | 84.6 |