Gemma 4 31B Instruct is Google DeepMind's 30.7B dense multimodal model supporting text and image input with text output. Features a 256K token context window, configurable thinking/reasoning mode, native function...
Model details →Google: Gemma 4 31B vs Mistral: Codestral 2508 (batch)
Google: Gemma 4 31B wins on 5 of 6 axes — pricing and capability skew in its favour for most workloads.
Mistral's cutting-edge language model for coding released end of July 2025. Codestral specializes in low-latency, high-frequency tasks such as fill-in-the-middle (FIM), code correction and test generation. [Blog Post](https://mistral.ai/ne…
Model details →Side-by-side comparison
| Capability | Google: Gemma 4 31B | Mistral: Codestral 2508 (batch) | Winner |
|---|---|---|---|
| Context window Maximum number of input tokens the model can attend to in a single request. | 262K | 256K | 🏆 Google: Gemma 4 31B |
| Input price (per 1M) Cost per million input tokens billed by the provider. | $0.09 | $0.30 | 🏆 Google: Gemma 4 31B |
| Output price (per 1M) Cost per million output tokens billed by the provider. | $0.34 | $0.90 | 🏆 Google: Gemma 4 31B |
| Tool / function calling First-class support for emitting structured tool calls. | Yes | Yes | Tie |
| Vision input Accepts image inputs alongside text. | Yes | — | 🏆 Google: Gemma 4 31B |
| Reasoning mode Internal chain-of-thought / extended-thinking support. | Yes | — | 🏆 Google: Gemma 4 31B |
Frequently asked questions
Is Google: Gemma 4 31B better than Mistral: Codestral 2508 (batch)?
Google: Gemma 4 31B wins on 5 of 6 axes — pricing and capability skew in its favour for most workloads.
What's the price difference between Google: Gemma 4 31B and Mistral: Codestral 2508 (batch)?
Input: $0.09 vs $0.30 per 1M tokens. Output: $0.34 vs $0.90 per 1M tokens.
What context windows do Google: Gemma 4 31B and Mistral: Codestral 2508 (batch) support?
Google: Gemma 4 31B supports up to 262K tokens. Mistral: Codestral 2508 (batch) supports up to 256K tokens.
Do both Google: Gemma 4 31B and Mistral: Codestral 2508 (batch) support tool calling?
Google: Gemma 4 31B: yes. Mistral: Codestral 2508 (batch): yes.