Codestral 2508
CodingToolsMistral · Released August 2025 · Knowledge cutoff March 2025
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)
It belongs to Mistral's catalog, which spans Medium (capability), Small and Ministral (efficiency) and dedicated code models, all with European data-residency roots.
Facts and pricing
- Provider
- Mistral
- Context window
- 256K tokens
- Input price
- 3.3 credits / 1M tokens ($0.30 raw)
- Output price
- 9.9 credits / 1M tokens ($0.90 raw)
- Vision (image input)
- No
- Tool calling
- Yes
- Extended reasoning
- No
Credits are what Zeplik bills: 1 credit = $0.10, computed from the raw provider rate with a 1.10x margin. Raw prices shown per 1M tokens.
Try Codestral 2508 now
What Codestral 2508 is best for
- Writing, reviewing and refactoring code; agentic coding workflows
- Long documents: the 256K-token window holds lengthy reports, contracts or papers whole
- Tool use and agents: reliably calls functions, so it can search, run skills and drive workflows
Example prompts
Prompts that suit a coding model like Codestral 2508:
- Refactor this function to be testable and write the tests
- Explain what this regex does and rewrite it to be readable
- Design a database schema for a multi-tenant invoicing app
- Review this pull request diff for correctness and edge cases
Mistral family
| Model | Context | Input cr/M | Output cr/M | Released |
|---|---|---|---|---|
| Mistral Medium 3.5 | 262K | 16.5 | 82.5 | April 2026 |
| Mistral Small 4 | 262K | 1.7 | 6.6 | March 2026 |
| Devstral 2 2512 | 262K | 4.4 | 22.0 | December 2025 |
| Ministral 3 14B 2512 | 262K | 2.2 | 2.2 | December 2025 |
| Ministral 3 8B 2512 | 262K | 1.7 | 1.7 | December 2025 |
| Ministral 3 3B 2512 | 131K | 1.1 | 1.1 | December 2025 |
| Mistral Large 3 2512 | 262K | 5.5 | 16.5 | December 2025 |
| Voxtral Small 24B 2507 | 32K | 1.1 | 3.3 | October 2025 |
| Mistral Medium 3.1 | 131K | 4.4 | 22.0 | August 2025 |
| Codestral 2508 | 256K | 3.3 | 9.9 | August 2025 |
Frequently asked questions
- What is Codestral 2508?
- Codestral 2508 is an AI model by Mistral. 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)
- How much does Codestral 2508 cost on Zeplik?
- Input tokens cost 3.3 credits per million and output tokens 9.9 credits per million (1 credit = $0.10; the raw provider rates are $0.30 and $0.90 per million). New accounts start with free credits.
- How long can a conversation with Codestral 2508 be?
- Codestral 2508 has a 256K-token context window (256,000 tokens), which covers the conversation plus any documents you attach.
- Does Codestral 2508 support images and tools?
- Codestral 2508 is text-only and supports tool calling.
Related models
Mistral Medium 3.5 is a dense 128B instruction-following model from Mistral AI. It supports text and image inputs with text output, and is designed for agentic workflows, coding, and complex...
Mistral Small 4 is the next major release in the Mistral Small family, unifying the capabilities of several flagship Mistral models into a single system. It combines strong reasoning from...
Devstral 2 is a state-of-the-art open-source model by Mistral AI specializing in agentic coding. It is a 123B-parameter dense transformer model supporting a 256K context window. Devstral 2 supports exploring...
The largest model in the Ministral 3 family, Ministral 3 14B offers frontier capabilities and performance comparable to its larger Mistral Small 3.2 24B counterpart. A powerful and efficient language...
A balanced model in the Ministral 3 family, Ministral 3 8B is a powerful, efficient tiny language model with vision capabilities.
The smallest model in the Ministral 3 family, Ministral 3 3B is a powerful, efficient tiny language model with vision capabilities.