Access, auth, and limits
Point your client at https://mcp.searchable.com/mcp and sign in. Searchable uses OAuth 2.1 with PKCE and Dynamic Client Registration, so the client registers itself, opens Searchable in your browser, and shows a consent screen listing what it will be allowed to do. No API key in a config file, and origin validation on the server side. Probed on 2026-07-30, an unauthenticated JSON-RPC initialize returns 401, so the endpoint is live and genuinely gated. An API-key connection is still documented as the legacy path if your client cannot do OAuth.
The plan rule is simple and worth stating plainly: MCP requires a paid Searchable plan, and both connection methods are unavailable on the free plan, as are API keys. Once you are paying, though, there is no further tier-gating, MCP is listed on Pro at $125/mo exactly as it is on Enterprise at $999/mo. Safety defaults are sensible: read-only unless you grant write scope, and the three write tools (generate_report, trigger_audit, refresh_sitemap) are the low-risk kind.
Where it works and what to build
Client coverage is broad because it is a remote streamable-HTTP server: Claude.ai on web and mobile as a connector, Claude Desktop and Claude Code, ChatGPT through developer-mode connectors, Cursor, Windsurf, and anything else that speaks MCP. Nothing to install per machine, and interactive cards render results inline in clients that support them.
Where it earns its keep is the question a dashboard makes tedious. Ask why your visibility on a topic dropped, and the agent can pull get_visibility_by_topic, check get_share_of_voice against competitors, look at which domains the models cited with search_sources, then cross-reference get_gsc_performance and get_ai_traffic to see whether real traffic moved with it. That chain is four dashboards and a spreadsheet by hand. With write scope granted it can finish by calling trigger_audit and generate_report so the follow-up work is queued and the summary is ready to send.
Which AI clients can use Searchable MCP?
Not every AI assistant supports MCP natively. Per-client picture for Searchable:
| Client | Support | Notes |
|---|---|---|
| Claude (Desktop/Web) | Via connectors | Added as a connector on Claude.ai (web and mobile) and Claude Desktop; OAuth sign-in, no key pasting. |
| Claude Code | Native MCP | Documented client for the mcp.searchable.com/mcp streamable HTTP server. |
| ChatGPT | Via connectors | Supported through ChatGPT developer-mode connectors. |
| Cursor | Native MCP | Standard MCP config against the same server URL. |
| Windsurf | Native MCP | Listed as supported alongside any other MCP-compatible client. |
How to connect Searchable 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": {
"searchable": {
"command": "npx",
"args": ["mcp-remote", "<YOUR_MCP_URL>"]
}
}
}
// HTTP-transport MCP. Get your <YOUR_MCP_URL> from the vendor dashboard.
// Server source: https://mcp.searchable.com/mcp
// Claude Code (one-liner): claude mcp add searchable <YOUR_MCP_URL> --transport httpSome MCP servers require an API key. Check the source page for setup: mcp.searchable.com/mcp.
Does Searchable MCP cost extra?
The Searchable MCP server is community-built — free to run, but you still pay for your Searchable subscription.
- Community MCPs are normally self-hosted from a GitHub repo. There's no extra license fee.
- Your Searchable account still needs whatever plan is required to access the data the MCP server queries.
- Trade-off: zero MCP cost, but no guarantee it'll keep working when Searchable changes their UI or API.
Searchable also publishes a REST API
Searchable publishes an API, but per-tier gating wasn't extracted automatically.
What you can do with Searchable 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 Searchable 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 Searchable account
- Multi-account workflows without context-switching the Searchable UI
- Reports you'd normally spend hours on, done before standup
Production agent workflows
- Embed in your internal Codex / Cursor agents that triage Searchable data daily
- Combine with other MCP servers in a multi-step agent chain
- Constrain agents to read-only roles to keep production Searchable safe
Limits and gotchas
- Official MCP server — published by Searchable themselves, so it should track product changes.
- Fallback: the REST API works for any flow the MCP server doesn't cover.
- Rate limits: MCP calls count against your underlying Searchable API quota. Burst usage from a curious agent can drain a daily allowance fast.
Agent-readiness verdict
A 9, and the reason Searchable is worth a look even next to cheaper trackers. The surface is broad enough that an agent can run a real investigation rather than fetch one metric, the auth is the modern pattern rather than a key in a JSON file, read-only-by-default with scoped writes is the right safety posture, and access is on the $125 entry plan instead of behind an enterprise upsell. What keeps it off a 10: no webhooks, so an agent cannot be woken by a change, and MCP needs a paid plan, so there is no free way to prototype against it.
Scored by Joonas (TMB).
Searchable MCP & API FAQ
Is the Searchable MCP server official?+
Can I use the Searchable MCP with Cursor or Codex?+
Sources
- Searchable official site: https://www.searchable.com
- MCP server page: https://mcp.searchable.com/mcp
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.