Devstral Small
At a glance
- Price per 1M tokens (input)
- $0.10
- Price per 1M tokens (output)
- $0.30
- Context
- 256,000 tokens
- Free access
- yes (see below)
Refreshed daily; data verified November 16, 2026. Published August 12, 2026, last price change September 18, 2026.
Devstral Small by Mistral AI is a compact language model built for developers and technical teams that need lightweight automation without heavy infrastructure. It suits tasks such as code assistance, structured data extraction, and agent workflows that require step-by-step reasoning. Tool calling, which lets the model trigger external functions or APIs, enables it to act on information rather than only generate text. Mistral's approach emphasizes efficient, open-weight models that can be self-hosted and integrated into existing stacks.
Specifications & pricing
| Input (per 1M tokens) | $0.10 |
|---|---|
| Output (per 1M tokens) | $0.30 |
| Cache read (per 1M tokens) | $0.01 |
| Context window | 256,000 tokens |
| Max output | 256,000 tokens |
| Capabilities | tool calling, reasoning |
LiteLLM community dataset (MIT), verified November 16, 2026. Official Mistral AI pricing.
What Devstral Small would cost on your workload — run it through the cost calculator →
Price history
| Date | Input / 1M | Output / 1M |
|---|---|---|
| August 9, 2026 (first record) | $0.10 | $0.30 |
| September 18, 2026 | $0.10 | $0.30 |
Where to try Devstral Small for free
- Mistral AI offers a free chat — Le Chat (free plan). A vendor's free chat may run a different model from the same family — the exact model is not guaranteed.
Frequently asked questions
What is this model good for?+
It is well suited for coding support, extracting structured information from unstructured text, and running multi-step reasoning tasks. Its tool calling capability allows it to connect to external services and perform actions within defined workflows.
How do I get started?+
Access the model through Mistral's platform or deploy the open weights in your own environment. You can then send prompts through an API or local inference setup and define the functions the model is allowed to call.
How does it differ from sibling models?+
Devstral Small is the compact variant in the Devstral line, optimized for lower resource usage and faster inference. Larger sibling models handle more complex reasoning but require greater compute capacity.
What are its limitations?+
As a smaller model, it may struggle with highly complex or ambiguous tasks compared to larger variants. Tool calling also depends on clearly defined function schemas and reliable external endpoints.
Compare with others

Org chart: how to move your company onto AI
A practical map: which company roles and processes AI agents can take over, where to start, and in what order to roll it out.