Cloudflare MCP Server

Official suite of remote MCP servers that let Claude, Cursor, and other agents read and control your Cloudflare account.

Freemium· Free: $0 USD · Team: $4 USD per user/month for the first 12 months* · Enterprise: $21 USD per user/month for the first 12 months*MCP Servers

Visit website →

Best for

Developers and platform teams who already build on Cloudflare Workers, R2, D1, or Cloudflare One and want their coding agents to inspect, deploy to, and troubleshoot that account directly from chat.

Skip if

Teams that do not use Cloudflare, or security-sensitive environments that cannot route agent traffic through Cloudflare-hosted MCP endpoints and cannot self-host the servers.

Cloudflare MCP Server is a family of 16 official, remote Model Context Protocol servers maintained by Cloudflare that expose the Cloudflare platform to any MCP-compatible AI client (Claude Desktop, Claude Code, Cursor, Windsurf, the Cloudflare AI Playground, OpenAI's Responses API, and others). Rather than shipping one monolithic server, Cloudflare has sliced its surface area into domain-specific endpoints so agents only get the tools they actually need. Each server is hosted by Cloudflare at a dedicated URL (for example https://docs.mcp.cloudflare.com/mcp, https://browser.mcp.cloudflare.com/mcp, https://observability.mcp.cloudflare.com/mcp) and authenticates via OAuth against your Cloudflare account or an API token scoped to the relevant permissions.

The catalogue spans developer platform tools (Workers Bindings, Workers Builds, Workers Observability, AI Gateway, AutoRAG, Browser Rendering), networking and analytics (DNS Analytics, Radar, Digital Experience Monitoring, Logpush), security and compliance (Audit Logs, Cloudflare One CASB), documentation search (Docs AI Search over developers.cloudflare.com), and a blog reader plus a minimal Demo Day server that shows how to build your own. In practice this means an agent can search the Cloudflare docs, deploy and tail logs from a Worker, run a headless browser to fetch a live page, query real-time internet traffic data from Radar, or inspect misconfigurations flagged by CASB without you leaving the chat window.

Typical workflows include using Claude to debug a failing Worker by streaming its logs, asking Cursor to add a KV or R2 binding to the right wrangler.toml, having an agent pull an audit log to answer a compliance question, or wiring AutoRAG into a RAG pipeline for grounded answers. Source is Apache 2.0 on GitHub, so teams can also self-host or fork any of the servers.

Editor's take

This is the reference implementation for how a hyperscaler should ship MCP: official, remote-hosted, split into focused servers, and open source under Apache 2.0. If your stack sits on Cloudflare, wiring Claude or Cursor to the Workers Observability and Bindings servers pays for itself in the first debug session. Just scope your API tokens tightly - these servers can change production config, not just read it.

— The AI Tool Bible editorial team

Pros

✅ Official, first-party servers maintained by Cloudflare with 4k+ GitHub stars and active development.
✅ Remote-hosted at stable URLs so most clients need only a URL and OAuth login, no local install.
✅ Split into 16 focused servers so agents get relevant tools without prompt bloat.
✅ Covers most of the Cloudflare platform: Workers, R2, KV, D1, Radar, DNS analytics, AI Gateway, AutoRAG, Browser Rendering, audit logs, CASB.
✅ Works with any MCP-capable client - Claude Desktop, Claude Code, Cursor, Windsurf, and OpenAI Responses API.
✅ Apache 2.0 source is on GitHub, so teams can self-host, audit, or extend any server.
✅ Docs AI Search server gives agents grounded answers over the full developers.cloudflare.com corpus.

Cons

⚠️ Only useful if you already run on Cloudflare - it is not a general cloud MCP.
⚠️ Advanced capabilities inherit Cloudflare paywalls (Browser Rendering, AutoRAG, CASB, Cloudflare One all require paid plans).
⚠️ Write-capable tools mean a compromised or over-permissioned token can push destructive changes; scoping API tokens carefully is on you.
⚠️ Because the servers are hosted by Cloudflare, your agent's queries and returned data traverse Cloudflare infrastructure - not ideal for teams that want everything to stay in-VPC.
⚠️ OAuth and remote-MCP support is still uneven across MCP clients, so setup varies and some clients need a manual token instead.
⚠️ Sixteen separate endpoints can be confusing - discovery of which server owns which capability is not always obvious from client-side.

Use cases

Debugging Workers with live logs from chatDeploying and configuring Workers bindingsGrounded Q&A over Cloudflare developer docsQuerying Radar for real-time internet traffic dataPulling audit logs for compliance answersRunning a headless browser via Browser RenderingAutoRAG pipelines against R2-hosted corporaReviewing Cloudflare One CASB findingsDNS analytics and Logpush inspectionBuilding custom MCP servers from the Demo Day template