Last updated: 2026-11-12
Production URL: depends on the region your Outset organization is hosted in. See Using Outset's MCP in our help center for the full list of endpoints. This guide uses the US endpoint, https://api.outset.ai/mcp/, in its examples.
<aside> 🌍
Not on our US deployment (e.g. EU-hosted)? Use your region's MCP URL from the help article wherever this guide says https://api.outset.ai/mcp/, including the resource parameter. Discover the OAuth endpoints from that URL (see OAuth below) rather than reusing the api.outset.ai ones. Our official connectors (e.g. Outset in Claude's connector directory) currently work only with the US deployment, so add Outset as a custom connector with your region's URL instead. Not sure which region you're on? Contact [email protected].
</aside>
GitHub: Outset-AI/outset-mcp — client manifests (Claude Code & Cursor plugins, Gemini CLI extension, official MCP registry) and quick-connect snippets
Audience: developers building agents that need to connect to Outset. If you're comfortable with MCP and OAuth Dynamic Client Registration, the TL;DR below is enough to get you wired up — the rest of this page is for everyone else.
POST https://api.outset.ai/mcp/ (US; other regions in the help article) — Streamable HTTP, JSON-RPC, stateless, single endpoint.mcp Python/TS SDK ≥ 1.27. No SSE, no WebSocket — every request stands on its own.https://api.outset.ai/.well-known/oauth-authorization-server — full RFC 8414 document with every endpoint URL, supported scopes, grant types, PKCE methods, and auth methods. Always fetch at runtime instead of hardcoding endpoint URLs. In any region, an unauthenticated call to the MCP endpoint returns a 401 whose WWW-Authenticate header names the RFC 9728 resource_metadata URL, which in turn lists that region's authorization server.https:// URL as your client_id, no registration round-trip. Open DCR per RFC 7591 still available. Loopback (http://127.0.0.1:*) or HTTPS redirect URIs only, either way.projects:read analytics:read studies:read. studies:write, analytics:write, and studies:launch (publish/unpublish + recruitment, may incur fees) are split out.resource= your region's MCP URL (https://api.outset.ai/mcp/ in the US) on /authorize and /token. Tokens are audience-bound to the MCP server.If you just want to try it: point npx @modelcontextprotocol/inspector at https://api.outset.ai/mcp/, click through the consent screen, and you're in.
The Model Context Protocol is an open spec from Anthropic for connecting LLM agents to tools and data. An MCP server exposes a list of tools (typed JSON-RPC methods) that any compliant MCP client — Claude Desktop, Cursor, MCP Inspector, your own agent built on the official Python/TypeScript SDK — can discover and call.
Outset's MCP server hosts the same study-creation, study-editing, and analytics tools that drive our internal AI co-editor and ATLAS analyst, exposed to outside agents. Concretely your agent can: