product updateMistral AI

Mistral adds workspace-level connector controls, multi-account authentication, and debugging tools

TL;DR

Mistral AI released new enterprise connector features including workspace-level access controls, multi-account authentication for single connectors, and a debugging tool for Model Context Protocol (MCP) connections. The updates address production deployment challenges for AI agents accessing enterprise data systems.

2 min read
0

Mistral adds workspace-level connector controls, multi-account authentication, and debugging tools

Mistral AI released new enterprise features for its connector system, addressing production deployment requirements for AI agents that access external data sources.

What's new

Workspace and org-level controls (Generally Available): Administrators can now set connector access per workspace and enable or disable individual tools within connectors. A finance workspace can access internal data sources while blocking open web access, while engineering teams get developer tools and internet connectivity from the same directory.

API keys with connector scopes (Generally Available): New API key types allow administrators to specify whether keys access workspace shared connectors, private connectors, or both. Combined with service accounts, this prevents automated workflows from impersonating the developers who created them.

Multi-account connectors (Generally Available): Users can authenticate multiple accounts to a single connector and switch between them per task. Each account maintains separate credentials and refresh tokens, eliminating the need to configure duplicate connectors for personal and work accounts.

Connectors Debugger (Public Preview): A diagnostic tool that tests MCP server connections through 11 validation steps, from initial server reach to MCP session establishment. The debugger identifies exact failure points, such as broken OAuth token exchanges, that previously required manual troubleshooting.

Connectors in Workflows (Public Preview): Workflow definitions can declare required connectors and execution accounts upfront. Studio resolves dependencies at runtime, preventing authentication failures during long-running or scheduled tasks.

Connectors in Vibe Code (Generally Available): Mistral's coding agent can now access local MCP servers and workspace-approved connectors using the same governance rules applied across the platform.

Integration coverage

Mistral's connector directory includes over 60 integrations across enterprise platforms:

  • Knowledge and data: Notion, Confluence, SharePoint, Google Drive, Databricks, Snowflake
  • Developer tools: GitHub, Netlify, Sentry
  • CRM and operations: Salesforce, HubSpot, ServiceNow, Linear, Asana
  • Communication: Gmail, Outlook, Slack
  • Payments: Stripe, PayPal, Square

Custom MCP connectors can be created for platforms not in the directory.

Availability

Enriched admin controls, API keys with connector scopes, multi-account connectors, and Connectors in Vibe Code are generally available. Connectors Debugger and Connectors in Workflows are in public preview. All features are live in Mistral Studio.

What this means

These updates target the gap between prototype AI agents and production deployments in enterprise environments. The workspace-level controls and scoped API keys address security and compliance requirements that prevent many organizations from deploying agents with data access at scale. Multi-account support and the debugging tool reduce operational friction for teams managing multiple identity contexts and troubleshooting connection failures. The focus on Model Context Protocol indicates Mistral's commitment to the emerging standard for AI-to-system connectivity.

Related Articles

product update

Safari 27 Adds Native MCP Server, Letting AI Agents Like Claude Code and Codex Control the Browser

Safari 27.0 introduces a local MCP server that lets AI coding agents control the browser to inspect DOM, network requests, and screenshots. Apple says the feature runs entirely on-device with no data sent to Apple.

product update

OpenAI Lists GPT-6 Luna Pro: A High-Reasoning Mode for Its Budget GPT-6 Model, Not a New Checkpoint

GPT-6 Luna Pro, listed on OpenRouter with a Sep 22, 2026 release date, is not a distinct model but GPT-6 Luna run with reasoning.mode set to 'pro' for higher-quality outputs on complex tasks. It carries a 1.1M token context window and costs $0.10 per 1M input tokens and $0.50 per 1M output tokens under standard routing.

product update

Anthropic Builds Physical Biology Lab Where Claude Directs Robots to Run Drug Experiments

Anthropic is setting up a physical biology lab in the San Francisco area where Claude will guide robots through drug experiments with minimal human involvement. The move follows the company's $400 million acquisition of Coefficient Bio and the addition of Novartis CEO Vas Narasimhan to its board.

product update

Meta's Muse AI Agent App Hits 730,000 Downloads, Overtakes ChatGPT on iOS Charts

Meta's Muse AI agent app overtook ChatGPT as the top free iOS app in the U.S., racking up 730,000 downloads in its first five days, according to Sensor Tower. The app, powered by Meta's Muse Spark model family, marks Zuckerberg's biggest push yet into AI agents.

Comments

Loading...