Toolofficial publisher

code_interpreter (Mistral)

Tool : A precise action the AI can call · requires mistral-api

Python sandbox for Mistral agents.

The European option for code execution.

Docs

Overall score

6.7 /10

Security: one caveat, see the card

Utility

7.0

Reliability

6.0

Security

One caveat

Ease

1 click · Comfortable

Security reflects where data goes, requested permissions and incidents on record at the verification date. It is not a guarantee of zero risk: check the deployment mode and the connections you enable.

What it does, concretely

When an agent needs to perform calculations, analyze data, or produce charts, it cannot do so from its language model alone. Mistral's code_interpreter tool solves this by allowing the agent to execute Python code in an isolated environment, directly on the server side.

In practice, the agent receives a request, generates Python code, sends it to the tool which runs it in a secure sandbox, then retrieves the result or the produced chart. All execution takes place within Mistral's internal infrastructure, without any code running on the user's machine.

This tool is aimed at developers building agents via the Mistral API who need to automate data analysis, visualization, or computation tasks. It is also accessible directly through the Conversations interface, without needing to create an agent beforehand.

However, code_interpreter is not available in the classic Chat Completions API. It is also not suitable for users who want to run code in their own local environment, as execution is entirely managed on Mistral's side.

What it offers

·Python code execution in an isolated server-side sandbox (paid).
·Chart and visualization generation from provided data (paid).
·Available as a built-in tool, ready to use without additional configuration (paid).
·Usable directly via the Conversations interface without prior agent creation (paid).
·Compatible with the Conversations API and the Agents API (paid).
·Can be combined with other built-in tools such as web search or image generation within the same agent (paid).

Our tips

You can activate multiple tools simultaneously within the same agent, for example code_interpreter and web_search, to combine information retrieval and data processing.
To use code_interpreter, specify it in the tools parameter when creating an agent or when making a direct call to the Conversations API.
If you want to test the tool quickly without creating an agent, use the Conversations interface, which supports built-in tools directly.
Do not attempt to use code_interpreter via the Chat Completions API, as it is not supported there; use exclusively the Conversations API or the Agents API.

Watch out

The tool is not available in the Chat Completions API, which may catch developers off guard if they are used to that standard endpoint.
Execution takes place entirely within Mistral's environment: it is not possible to access local resources or libraries not available in the sandbox.

In practice

Price: Paid · from 28 €
Commercial use: Commercial use: yes
Available on: api
Host: mistral-api
FR interface: No

Security and data

Rating: Security: one caveat, see the card
Capabilities: readswritesexecutesnetwork
Your data: Data with the vendor
Maintenance: Active

Links and repository checked on 25 October 2026 · added on 1 May 2025

How to install code_interpreter (Mistral)

1. Create an account and obtain an API key on the Mistral platform. 2. Install the Mistral Python client in your development environment. 3. When creating an agent via client.beta.agents.create(), add code_interpreter to the tools parameter. 4. You can also call the tool directly via the Conversations API without creating an agent. 5. Never include your API key in shared or public code; store it in an environment variable.

tools=[{"type":"code_interpreter"}]

Official documentation ↗

In the same category

Whole category Code →