Codestral Embed

GA

Premier code-embedding model (v25.05). Embeddings only (input-priced): $0.15/M input tokens (mistral.ai/pricing, 2026-06-28). Max input 8,192 tokens (contextLength '8k' in the open-source docs schema codestral-embed-25-05.ts). Output dimension configurable via output_dimension: default 1536, max 3072 (docs.mistral.ai/capabilities/embeddings/code_embeddings). releaseDate 2025-05-28 per the docs schema.

Codestral Embed by Mistral is tracked live in our LLM catalog, with a 8K (8192-token) context window. It is generally available (GA).

Last verified: 13 Aug 2026 · sourced from official provider documentation

Provider
Mistral
Status
GA
Input price
$0.15 / 1M tokens
Output price
Cached input
Blended price
Context window
8192 tokens (8K)
Max output
Embedding dimensions
1536-dim output vectors
Modality
text, code
Open weights
No — API only
Knowledge cutoff
Released
28 May 2025
API string
codestral-embed-2505

Source: Mistral official documentation ↗

FAQ

Codestral Embed — questions & answers

What is the context window of Codestral Embed?

Codestral Embed has a 8192-token context window (8K).

Is Codestral Embed deprecated?

No — Codestral Embed is generally available (GA) and not currently scheduled for deprecation or retirement in our tracker.

Does Codestral Embed support image or vision input?

Our catalog lists Codestral Embed's modalities as text, code; image/vision input is not among them.

Can I self-host Codestral Embed?

No — Codestral Embed is proprietary. Its weights are not publicly released, so it is available only through Mistral's API (or hosted partners), not for self-hosting.

What is the API model string for Codestral Embed?

The API model identifier for Codestral Embed is "codestral-embed-2505" when calling Mistral's API.