Codestral 2508
Mistral AI · released Aug 1, 2025
- Type
- Proprietary
- Input
- $0.30
- Output
- $0.90
- Cached
- $0.030
List price · per 1M tokens · Mistral AI at 256K context · source ↗
Our take
Written Aug 2, 2026Codestral 2508 is Mistral AI's hosted-only coding specialist, built for long-context code analysis with a 256,000-token request limit. It is a niche pick for teams already in the Mistral ecosystem who need to feed large codebases into a single prompt, though no independent quality scores exist to validate its performance.
Choose this when you need to analyse or generate code with up to 256,000 tokens of source context in one request, and you are comfortable with Mistral's own hosting at 13 tokens per second. Skip it if you need measured benchmark scores to justify the choice, if you want to self-host or avoid proprietary lock-in, or if throughput speed is critical to your workflow.
The case for it
- Among the longest request limits we hold for a coding model, at 256,000 tokens.
- Identical pricing across both tracked providers, so there is no need to shop for a cheaper host.
The case against it
- No benchmark scores in our data, so coding performance is unverified.
- 13 tokens per second on the primary host — with no comparison figure, we cannot say whether this is competitive.
- Proprietary weights with no disclosed licence, and no open alternative available.
How good is it?
We hold no score for this model.
We look for every model we track on every board we watch, and none of them has turned up Codestral 2508 — so there is no intelligence, coding, agentic or writing score to show you, not a low one, none.
Or rent it from someone else
Why this differs from the header. The strip above quotes Mistral AI's own list price. This is the cheapest live offer at the widest standard context we hold, whoever is serving it — a reseller undercutting a lab is ordinary commerce, not an error.
- per 1M tokens
- $0.30 in / $0.90 out
- Context served
- 256K
- Throughput
- Not measured
| Provider | In / out per 1M tokens | Context | Throughput | Trains on prompts | Logs prompts | Zero retention |
|---|---|---|---|---|---|---|
| OpenRouter | $0.30 / $0.90 | 256K | not measured | Unknown | Unknown | Unknown |
| Mistral AI | $0.30 / $0.90 | 256K | 57 tok/s | No | Yes30 days | Unknown |
Mistral AIAcross the 2 listings we hold: 1 say they do not train on prompts, 0 say they do and 1 do not say. 0 appear in the zero-retention registry we check; the rest are unknown to us rather than confirmed either way.
What each host's API supports
From the parameter list each endpoint publishes. Streaming is omitted: nothing we hold reports it, for any model.
- ✓
- Supported
- ✗
- Not supported
- Not published
- host gave no parameter list
| Provider | Tool calling | JSON output | Strict schema |
|---|---|---|---|
| OpenRouter | ✓ | ✓ | ✓ |
| Mistral AI | ✓ | ✓ | ✓ |
Tool calling: 2 of 2 listings say yes. JSON output: 2 of 2 listings say yes. Strict schema: 2 of 2 listings say yes.
When we formed this view
Dates behind this page
Prices last checked 9d ago
What we do not know about this model yet
- No board we watch has turned up a score, so we hold no quality figures at all.
- Nothing we hold says whether an endpoint streams, so we do not show it either way.
- 1 of 2 listings do not say whether they train on prompts.
Licence and identifiers
What the licence allowsWe hold no licence record for this model. Inside are the identifiers you need to pull it — its Hugging Face repo where we have one, our slug and a machine-readable card.
Licence
Commercial API terms. We hold no licence record for this model, so there is nothing to summarise here.
Identifiers
- Modality record
- text+file->text
- Catalogue slug
- mistralai-codestral-2508