Mistral Small
Mistral AI
At a glance
- Price per 1M tokens (input)
- $0.10
- Price per 1M tokens (output)
- $0.30
- Context
- 32,000 tokens
- Free access
- yes (see below)
Refreshed daily; data verified August 15, 2026. Published August 12, 2026.
Mistral Small is a compact, efficient model from Mistral AI designed for developers and businesses that need reliable language processing without the overhead of larger systems. It is well-suited for tasks like summarization, classification, and structured data extraction, where speed and cost-effectiveness matter more than maximum creative output. Its standout feature is native tool calling, which lets the model invoke external functions or APIs to complete workflows, making it a practical choice for automation and agentic applications. Mistral AI's approach emphasizes transparency and efficiency, offering models that balance performance with operational simplicity, and this small variant targets teams that want solid results with minimal infrastructure demands.
Specifications & pricing
| Input (per 1M tokens) | $0.10 |
|---|---|
| Output (per 1M tokens) | $0.30 |
| Context window | 32,000 tokens |
| Max output | 8,191 tokens |
| Capabilities | tool calling |
LiteLLM community dataset (MIT), verified August 15, 2026. Official Mistral AI pricing.
What Mistral Small would cost on your workload — run it through the cost calculator →
Where to try Mistral 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 in a business context?+
It excels at high-volume, structured tasks such as parsing customer feedback, generating concise internal reports, or powering chatbots that need to look up order statuses via an API. Because it supports tool calling, it can also orchestrate multi-step processes like booking a meeting or updating a CRM record, making it a fit for workflow automation.
How do I get started with using it?+
You can access it through Mistral AI's platform or via compatible API endpoints. The quickest path is to send a prompt that includes a description of the tool you want the model to call, along with the function's schema. Most frameworks, like LangChain or OpenAI-compatible clients, support it with minimal configuration, so you can test a basic call within an hour.
How does it differ from other models in the Mistral line?+
The Mistral line spans different sizes and capabilities. This small variant is optimized for lower latency and lighter compute, making it cheaper to run at scale, but it has less depth in reasoning and creative writing compared to larger siblings. It is the pragmatic choice when you need reliable function execution and straightforward text tasks, not when you need complex multi-turn debate or long-form narrative.
What are its main limitations I should plan for?+
It can struggle with nuanced instruction following or ambiguous requests, so you should write very explicit prompts. Its tool calling works best with well-defined schemas; if your API has irregular parameters, you may need to preprocess inputs. Also, it has a shorter memory for conversation context, so for long dialogues you must summarize or trim history, which can add engineering effort.
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.