Mistral AI has unveiled its new Agents API, a framework meant to turn language models into hands-on problem solvers for businesses. The Agents API lets AI agents handle tasks on their own, work together in teams, and plug into outside systems.
The new API builds on basic language models by adding tools for action, context tracking, and coordinating multiple specialized agents. Each agent can be hooked up with connectors and MCP tools to run Python code, search the web, create images powered by Black Forest Lab FLUX1.1 [pro] Ultra, or pull documents from Mistral Cloud. Connections to outside services are handled through the Model Context Protocol (MCP), Anthropic’s standard for linking APIs, databases, and user data.
Agents that remember and collaborate
Unlike regular chatbots, conversations with the Agents API are stateful—the context sticks around even when an interaction is paused or rolled back. Developers can start a conversation with a specific agent or connect directly to a model using connectors. Outputs stream in real time.
One of the main features is agent orchestration: multiple agents can work together on a single task, dividing up the work as needed. For example, a financial agent can hand off tasks to a web search agent or a computational agent. Each agent stays focused on its own area of expertise.
Mistral demonstrates the API in several demo applications. One example is a coding assistant that connects to GitHub, oversees a developer agent, and automates parts of the software development process.
Another demo shows an intelligent task coordination assistant using multi-server MCP architecture to transform call transcripts to PRDs and actionable Linear tickets, and tracks project deliverables.
A financial agent orchestrates multiple MCP servers to source financial metrics, compile insights, and archive results securely.
Other demonstrations include a digital nutritionist designed to help users establish goals, log meals, receive personalized suggestions, track daily achievements, and discover dining options that align with their nutritional targets.
The Agents API is designed to work alongside Mistral's Chat Completion API and provides a framework for building agent-based enterprise applications. Developers can define multiple agents, combine them, and link them into custom workflows.
