---
id: mistralai-workflows
version: "3.11.0"
license: Apache-2.0
license_treatment: permissive
maintenance: active
---

# mistralai-workflows — Mistral Workflows - Build reliable AI workflows with Python

License: permissive · Maintenance: active · Downloads: 196.5K/mo

## What it is and what it does

Mistral Workflows is a Python framework for defining and executing reliable, distributed AI workflows using decorators and async/await syntax. It wraps Mistral AI LLM calls and custom activities (tasks) into workflows with automatic retry logic, timeouts, and exactly-once execution semantics. The SDK is built on temporalio for durability and state management, integrates the opentelemetry stack for observability, and uses Pydantic for type validation.

You define activities (individual tasks) and workflows (orchestrations of activities) using Python decorators, then execute them with built-in fault tolerance and distributed scaling. It's aimed at production AI applications that need reliability guarantees—retries on failure, timeout enforcement, and structured logging—without manually implementing those patterns.

Use it for:

- Build multi-step LLM pipelines (e.g., research → analysis → report generation) with automatic retries and timeout protection.
- Orchestrate long-running AI tasks across multiple workers with exactly-once semantics and durable state.
- Integrate Mistral AI calls with external APIs and custom Python logic in a single declarative workflow.
- Monitor and trace distributed AI workflows using built-in OpenTelemetry instrumentation.
- Deploy fault-tolerant AI agents that survive worker failures and resume from checkpoints.

## Worth the install?

AI-flagged interpretation of the facts above — verify before relying.

Mistral Workflows is a Python SDK for building AI-powered workflows with built-in fault tolerance, distributed execution, and observability using decorators and Pydantic validation.

Yes, if you are building production AI workflows with Mistral AI and need built-in reliability, distributed execution, and observability. The low install friction and active maintenance are strong signals. However, the Beta status, Python 3.12+ requirement, and temporalio dependency complexity mean you should verify that the maturity level and backend setup align with your deployment model before committing to a large workflow system.

## Install

pip install mistralai-workflows
uv add mistralai-workflows
poetry add mistralai-workflows

## Installing mistralai-workflows

Before you install: Low install friction with a pure-wheel distribution. Active maintenance (released 3 days ago). Requires Python 3.12–3.14 and pulls in 17 runtime dependencies including temporalio, httpx, and observability libraries (opentelemetry stack), which adds moderate complexity to your environment.

License in practice: Apache-2.0 permissive license allows commercial use, modification, and distribution with minimal restrictions—standard for open-source infrastructure.

Quickstart:

pip install mistralai-workflows

from mistralai.workflows import workflow, activity
from datetime import timedelta

@activity
async def get_weather(city: str) -> str:
    return f"Weather in {city}: Sunny"

@workflow.define
class WeatherWorkflow:
    @workflow.run
    async def run(self, city: str) -> str:
        weather = await workflow.execute_activity(
            get_weather, city, start_to_close_timeout=timedelta(seconds=10)
        )
        return weather

Requires Python 3.12 or later; temporalio dependency may require additional system setup for workflow persistence and distributed coordination.

Verify before relying:

- Whether temporalio backend requires external service deployment or can run in-process for development.
- Performance characteristics and scalability limits for concurrent workflows.
- Maturity and stability guarantees given Beta development status.

## Package facts

- License: Apache-2.0 (permissive)
- Python support: supports_current
- Install friction: low
- Maintenance: active
- Downloads: 196.5K/month (top 15,000 on PyPI)
- Known vulnerabilities: none known

## Tags

ai workflow orchestration python, mistral llm workflow engine, distributed task execution framework, fault-tolerant workflow automation, temporal-based workflow sdk, llm pipeline orchestration, reliable ai task scheduling, workflow-orchestration, llm-integration, distributed-execution

[View on SkillFed](https://skillfed.io/packages/mistralai-workflows) · [View on PyPI](https://pypi.org/project/mistralai-workflows/)