Mistral's dense model for agents that explore codebases and edit multiple files at once. Runs on a single H200 or a pair of A100/H100.
| SWE-bench Verified | 72.2 | |
|---|---|---|
| SWE-bench Multilingual | 61.3 | |
| Terminal-Bench 2 | 32.6 |
Which GPU does Devstral 2 123B need?
The entry configuration is 1x H200, which delivers around 14 tokens per second. For more parallel requests and headroom on speed, go with 2x H200.
How much does Devstral 2 123B cost?
An hour of Devstral 2 123B runs from $3.86 to $8.28 depending on the GPU configuration. Billing is hourly: you pay for server time, not for tokens.
How do I call Devstral 2 123B from my code?
Once deployed you get an OpenAI-compatible endpoint: point any OpenAI client at the base_url you receive and pass devstral-2:123b as the model name. Code written against the OpenAI API needs no changes.
Does the data stay private?
Yes. Devstral 2 123B runs on a dedicated GPU server inside a private environment: requests and responses never reach the model vendor and are not used for training.
What context window and license does Devstral 2 123B have?
Devstral 2 123B holds 32K tokens of context per request and ships under the Modified MIT license, so you can use it commercially on your own hosting terms.