Devstral Medium

Devstral Medium is a legacy model in the devstral line, retained for compatibility with older integrations. It is the first version of its line in the registry, released on 2025-07-10. The model has a context window of 128000 tokens and a maximum output of 128000 tokens. It accepts text and returns text. It supports tool calling but has no reasoning mode. The weights are open. It is served by 2 hosts. Pricing starts at $0.40 per million input tokens and $2.00 per million output tokens. Its knowledge cutoff is 2025-05.

First version of its line in the registry.

• ·128k token context window
• ·128k token max output
• ·Supports tool calling
• ·Open weights for self-hosting
• ·Text-to-text processing

• Reach for it for maintaining legacy integrations
• Reach for it for text-based tool calling workflows
• Reach for it for tasks needing a 128k context window

2 hosts serve this model at the price above · the maker's documentation

The registry keeps the six most recent versions of a line. Older ones stay in the file and can be surfaced without a rebuild.

What is the context window size?

The context window is 128000 tokens, and the maximum output is also 128000 tokens.

Does it support reasoning?

No, it does not have a reasoning mode. It supports tool calling and processes text only.

Are the weights open?

Yes, the weights are open. The model is served by 2 hosts and pricing starts at $0.40 per million input tokens and $2.00 per million output tokens.