Codestral 2508
MistralSpecialized
Tool UseStructured Output
About this model
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/news/codestral-25-08)
Performance Tier
Specialized
Codestral 2508 is a specialized model from Mistral : built for a specific domain.
Domain-specific model. Optimized for a particular task such as code generation, image creation, or web search.
Pricing
This model is included in Elosia plans| Type | per 1M tokens |
|---|---|
| Input (prompt) | $0.300 |
| Output (completion) | $0.900 |
| Cache read | $0.030 |
Capabilities
Context Length256K
Max Output Tokens—
TokenizerMistral
Inputtext
Outputtext
Release DateAugust 1, 2025
Benchmarks
Programming
HumanEval
90.5%
SWE-bench Verified
52%
Recommended Use Cases
Coding
Strengths
- Purpose-built for code generation by Mistral
- Strong HumanEval performance (90.5%)
- Supports 80+ programming languages
- EU-based — GDPR-friendly code model
Limitations
- Specialized for code — not suited for general conversation
- Software engineering tasks trail top models (SWE-bench 52%)
Resources
This model may use your data for training
