One agent session
1M input tokens and 100K output tokens, no caching.
- Input1M x $0.30 = $0.30
- Output0.1M x $0.90 = $0.09
$0.39per session
Mistral AI - Closed weights
Codestral 25.08 pricing and specs
Codestral 25.08 is a closed-weights model from Mistral AI that costs $0.30 per 1M input tokens and $0.90 per 1M output tokens on the first-party API, with a 128K-token context window.
Last verified from docs.mistral.ai. All sources
- Input / 1M
- $0.30
- Output / 1M
- $0.90
- Context
- 128K tokens
- Weights
- Closed
Pricing per 1M tokens
| Item | Price / 1M tokens |
|---|---|
| Input | $0.30 |
| Output | $0.90 |
| Cache read | $0.03 |
| Batch inputBatch requests | $0.15 |
| Batch outputBatch requests | $0.45 |
Note: This is a code completion (FIM) model, listed under Code models on the pricing page. Batch prices come from the pricing page's Batch mode, which applies a factor of 0.5 to standard prices. Priority Tier adds 75% and regional inference adds 10%. The docs give the context only as '128k', and the Vercel AI Gateway lists 128000. No primary page states a max output. The models overview labels it 'Premier' and the model card has no weights tab, so weights are recorded as closed. The model card is dated July 30, 2025; release date uses the changelog date (July 29).
Specs
- Context window
- 128,000 tokens
- Release date
- Provider
- Mistral AI
- API model id
codestral-2508- Vercel AI Gateway id
mistral/codestral- Weights
- Closed weights
What Codestral 25.08 costs in practice
Worked from the standard list rates above, with the math shown so you can swap in your own token counts.
A chat app
10,000 requests a day at 2K input and 500 output tokens each.
- Input20M x $0.30 = $6.00
- Output5M x $0.90 = $4.50
$10.50per day
$315.00 per 30-day month
The same session at a 90% cache hit
90% of the input is read from the prompt cache. The one-time cache write premium is not included.
- Cached input0.9M x $0.03 = $0.027
- Uncached input0.1M x $0.30 = $0.03
- Output0.1M x $0.90 = $0.09
$0.147per session
Compare with
| Model | Input / 1M | Output / 1M |
|---|---|---|
| Codestral 25.08 (this page) | $0.30 | $0.90 |
| GLM-4.6VZ.ai | $0.30 | $0.90 |
| MiMo-V2.6-ProXiaomi MiMo | $0.435 | $0.87 |
| Mercury 2Inception | $0.25 | $0.75 |
| Mercury 2.5Inception | $0.20 | $0.75 |
| Mistral Small 4Mistral AI | $0.15 | $0.60 |
| Mistral Large 3Mistral AI | $0.50 | $1.50 |
| Ministral 3 14BMistral AI | $0.20 | $0.20 |
| Ministral 3 8BMistral AI | $0.15 | $0.15 |
| Ministral 3 3BMistral AI | $0.10 | $0.10 |
| Mistral Large 4Mistral AI | $1.36 | $4.18 |
FAQ
How much does Codestral 25.08 cost per million tokens?
Codestral 25.08 costs $0.30 per 1M input tokens and $0.90 per 1M output tokens on Mistral AI's first-party API, per docs.mistral.ai on October 17, 2026. Per 1M tokens, cache reads cost $0.03, and the batch rate is $0.15 input and $0.45 output.
What is Codestral 25.08's context window?
Codestral 25.08 has a 128,000-token context window.
Are Codestral 25.08's weights open?
No. Codestral 25.08 is a closed-weights model. You call it through Mistral AI's API as codestral-2508 or through Vercel AI Gateway as mistral/codestral.
When was Codestral 25.08 released?
Mistral AI released Codestral 25.08 on July 29, 2025.
Sources
Last verified from docs.mistral.ai.
Newsletter
Get model price changes as they ship
One email per week. New model releases, price cuts and what they mean for your bill.