Connect the dots: Build with built-in and custom MCPs in Studio

Mistral has released Connectors in Studio to help developers build AI applications that rely on enterprise data. All built‑in connectors and custom MCPs are now accessible through API/SDK for use with model and agent calls. The update introduces direct tool calling, giving developers precise control over tool invocation without authentication barriers, and human‑in‑the‑loop approval flows that allow secure review before tool execution. Programmatic access is provided for creating, modifying, listing, deleting connectors, listing their tools, and running them directly. Connectors are centrally registered, making them available across Mistral apps such as LeChat and AI Studio, with Vibe coming soon. They can be used via the Conversation API, Completions API, and Agent SDK to facilitate complex workflows and integration with enterprise systems like CRMs, knowledge bases, and productivity tools. The article explains that building enterprise AI agents is easier, but the surrounding tasks—finding API docs, writing tool functions, setting up OAuth, handling token refresh, debugging pagination—often lead to duplicated integration layers. A connector packages an integration into a reusable entity using the MCP protocol, reducing security risks, improving observability, and eliminating duplication. Custom MCP connectors, once registered, become native tools for any conversation, agent, or workflow without rewriting logic or re‑implementing authentication. The text also describes how connectors can expose many tools, with tool_configuration controls to limit availability, and how connectors can be called directly for deterministic workflows. The release is available in Studio’s public preview, and developers can start building today via the Studio console.