Codestral 25.08
Language modelCard summarizedOur cutting-edge language model for code completion released end of July 2025, Codestral specializes in low-latency, high-frequency tasks such as fill-in-the-middle (FIM) and code generation.
Active (GA)● Re-checked 12 Oct 2026 · 30 facts · 0 correctionsSource: Codestral 25.08 launch post ("Codestral 25.08 and the complete Mistral coding stack") and model docs page; no model card published
Use it via API
Where to call it
- Mistral API
codestral-2508
Good for
- Inline code completion and multi-line edits (FIM) in IDEs, including Mistral Code for JetBrains and VS Code
- Latency-sensitive, high-frequency code generation
- Short- and long-context FIM completion
Overview
Codestral 25.08
Our cutting-edge language model for code completion released end of July 2025, Codestral specializes in low-latency, high-frequency tasks such as fill-in-the-middle (FIM) and code generation.
- Cost
- $0.30 in
- Context
- 128K
- Input / output
- $0.30 / $0.90
- Max output
- —
- Knows up to
- —
- Released
- 30 Jul 2025
Codestral 25.08 is Mistral's update to its fill-in-the-middle code-completion model, the base layer of its enterprise coding stack (with Codestral Embed, Devstral and the Mistral Code IDE extension). Mistral reports 30% more accepted completions, 10% more retained code and 50% fewer runaway generations than earlier versions, measured in live IDE use. It is a Premier (closed-weights) model deployable in the cloud, a VPC or on-prem. launch post
codestral-2508- Input / output per 1M
- $0.30 / $0.90
- Context window
- 128K
- Max output
- —
- Knows up to
- —
- Reasoning
- —
- Weights
- Closed
About this card. A model card is the maker's own account of a model: what it's for, how it was trained and tested, and where it falls short. This page condenses Mistral AI's Codestral 25.08 launch post ("Codestral 25.08 and the complete Mistral coding stack") and model docs page; no model card published. Every claim carries its source page.