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 →Mistral: Codestral 2508 vs Qwen: Qwen3 30B A3B Instruct 2507
Qwen: Qwen3 30B A3B Instruct 2507 wins on 3 of 6 axes — pricing and capability skew in its favour for most workloads.
Qwen3-30B-A3B-Instruct-2507 is a 30.5B-parameter mixture-of-experts language model from Qwen, with 3.3B active parameters per inference. It operates in non-thinking mode and is designed for high-quality instruction following, multilingual u…
Model details →Side-by-side comparison
| Capability | Mistral: Codestral 2508 | Qwen: Qwen3 30B A3B Instruct 2507 | Winner |
|---|---|---|---|
| Context window Maximum number of input tokens the model can attend to in a single request. | 256K | 262K | 🏆 Qwen: Qwen3 30B A3B Instruct 2507 |
| Input price (per 1M) Cost per million input tokens billed by the provider. | $0.30 | $0.09 | 🏆 Qwen: Qwen3 30B A3B Instruct 2507 |
| Output price (per 1M) Cost per million output tokens billed by the provider. | $0.90 | $0.30 | 🏆 Qwen: Qwen3 30B A3B Instruct 2507 |
| Tool / function calling First-class support for emitting structured tool calls. | Yes | Yes | Tie |
| Vision input Accepts image inputs alongside text. | — | — | Tie |
| Reasoning mode Internal chain-of-thought / extended-thinking support. | — | — | Tie |
Frequently asked questions
Is Mistral: Codestral 2508 better than Qwen: Qwen3 30B A3B Instruct 2507?
Qwen: Qwen3 30B A3B Instruct 2507 wins on 3 of 6 axes — pricing and capability skew in its favour for most workloads.
What's the price difference between Mistral: Codestral 2508 and Qwen: Qwen3 30B A3B Instruct 2507?
Input: $0.30 vs $0.09 per 1M tokens. Output: $0.90 vs $0.30 per 1M tokens.
What context windows do Mistral: Codestral 2508 and Qwen: Qwen3 30B A3B Instruct 2507 support?
Mistral: Codestral 2508 supports up to 256K tokens. Qwen: Qwen3 30B A3B Instruct 2507 supports up to 262K tokens.
Do both Mistral: Codestral 2508 and Qwen: Qwen3 30B A3B Instruct 2507 support tool calling?
Mistral: Codestral 2508: yes. Qwen: Qwen3 30B A3B Instruct 2507: yes.