Access, auth, and limits
OAuth 2.0 against auth.tryprofound.com, with bearer tokens carried in the header. Running a separate authorization host rather than serving tokens from the resource server is the more careful pattern, and it suggests the MCP surface was designed rather than retrofitted.
The gap is everything downstream of the handshake. There is no public page listing tools, no statement of which plans include MCP access, and no setup guide, so both the capability and the entitlement are questions for Profound's sales team rather than things you can check yourself.
How to connect Profound MCP to Claude Desktop
Add this block to ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows), then restart Claude Desktop. Cursor and Codex use a similar schema.
{
"mcpServers": {
"profound": {
"command": "npx",
"args": ["mcp-remote", "<YOUR_MCP_URL>"]
}
}
}
// HTTP-transport MCP. Get your <YOUR_MCP_URL> from the vendor dashboard.
// Server source: https://mcp.tryprofound.com/mcp
// Claude Code (one-liner): claude mcp add profound <YOUR_MCP_URL> --transport httpSome MCP servers require an API key. Check the source page for setup: mcp.tryprofound.com/mcp.
Does Profound MCP cost extra?
Profound ships an official MCP server that uses your existing subscription credentials.
- No separate MCP licence to buy. You pay for your Profound plan, the MCP server queries that account.
- Whether it's gated to a specific tier wasn't extracted automatically — verify the access requirements on the MCP page or pricing docs.
Profound also publishes a REST API
Profound publishes an API, but per-tier gating wasn't extracted automatically.
What you can do with Profound from an AI agent
Three shapes of agent workflow. Pick the row that matches how you actually use Claude / Cursor / Codex day to day.
Skip the dashboard, ask Claude
- "What keywords moved this week?" answered in seconds
- Quick one-off questions without leaving the editor
- Pipe Profound data into your personal notes / docs without writing a script
Run client analysis at agent speed
- Hand the MCP server to your team — every consultant can ask Claude about any client's Profound account
- Multi-account workflows without context-switching the Profound UI
- Reports you'd normally spend hours on, done before standup
Production agent workflows
- Embed in your internal Codex / Cursor agents that triage Profound data daily
- Combine with other MCP servers in a multi-step agent chain
- Constrain agents to read-only roles to keep production Profound safe
Limits and gotchas
- Official MCP server — published by Profound themselves, so it should track product changes.
- Fallback: the REST API works for any flow the MCP server doesn't cover.
- Auth flow: MCP servers usually inherit the underlying API auth method (oauth2). Expect to configure that on first install.
- Rate limits: MCP calls count against your underlying Profound API quota. Burst usage from a curious agent can drain a daily allowance fast.
Agent-readiness verdict
The auth layer is properly built, with a dedicated authorization server rather than a bolted-on token check. The documentation is the gap: nothing about tools or tier gating is public, so evaluating it before you buy is not really possible.
Strong agent-fit. Profound has an official MCP server and documented public APIs with Node and Python SDKs, enabling seamless integration with Claude and ChatGPT workflows. The main limitation is lack of OpenAPI specs and structured outputs, which can require additional parsing work in agentic chains.
Scored by Joonas (TMB).
Profound MCP & API FAQ
Is the Profound MCP server official?+
Can I use the Profound MCP with Cursor or Codex?+
Does Profound have a public API?+
Sources
- Profound official site: https://www.tryprofound.com
- MCP server page: https://mcp.tryprofound.com/mcp
- API docs: https://docs.tryprofound.com/introduction
Data verified by Joonas on the dates shown. MCP server status auto-rechecked weekly.
10+ years in digital marketing. I review marketing software for AI-stack fit: real pricing, MCP and API support, and how cleanly each tool drops into an AI agent workflow, cross-checked against verified data and real user feedback.