How to Use MCP Servers in ChatGPT (Developer Mode + Apps SDK)
ChatGPT supports MCP now — the same open protocol Anthropic introduced in late 2024 — which means the same servers that power Claude can power ChatGPT. This guide walks through enabling Developer Mode, connecting a custom MCP server, and using Apps SDK integrations, with Carly as the working example. If you’ve been waiting for ChatGPT to catch up on tool use, this is it.
What ChatGPT’s MCP support actually does
MCP is a standard for letting an AI client call external tools — calendar, CRM, GitHub, your internal database, anything. ChatGPT can now connect to MCP servers and call those tools mid-conversation, which is what Claude has been doing in Claude Desktop and Claude Code for over a year. The one meaningful difference: ChatGPT only supports remote MCP servers — HTTPS endpoints you connect to over the network. It doesn’t support local stdio servers (the node my-server.js kind that Claude Desktop and Cursor run on your laptop). So you either need a hosted MCP endpoint, or you need to expose a local server as a remote one with a bridge like mcp-remote.
Plan availability
Custom MCP connectors live behind ChatGPT’s Developer Mode, which is available on Plus, Pro, Team, Enterprise, and Edu plans. Free ChatGPT does not support custom connectors. Apps SDK is the separate, developer-facing path: it’s how teams build integrations that show up in ChatGPT’s official app catalog for any user to install — no Developer Mode required on the user side. If you’re a builder, OpenAI’s MCP docs cover both paths and the Apps SDK examples repo has working code.
Two ways to connect
There are two flavors of “MCP in ChatGPT” and they get conflated all the time:
- Custom Connectors / Developer Mode — for individuals adding an MCP server to their own ChatGPT account. You paste in a URL, OAuth or paste a token, and the tools show up in your chats. Good for personal workflows and trying servers nobody has packaged yet.
- Apps SDK — for developers publishing an MCP-backed integration that any ChatGPT user can install from the app catalog. Apps SDK uses MCP under the hood but adds UI components, OAuth scaffolding, and review/distribution.
We’ll walk through Custom Connectors first because that’s what most people are searching for.
Step-by-step: enable Developer Mode in ChatGPT
- Open ChatGPT (web or desktop).
- Click your profile picture → Settings.
- Go to Connectors.
- Click Advanced at the bottom.
- Toggle Developer mode on.
You’ll get a warning that custom connectors run third-party code on your behalf. Read it, accept it, and you’re in. A new Add custom connector button appears in the Connectors panel.
If you don’t see the Advanced section, double-check you’re on Plus or higher and that your workspace admin (on Team/Enterprise) has allowed Developer Mode.
Step-by-step: add Carly as a custom MCP connector
Carly is an AI executive assistant. Its MCP server scopes specifically to booking pages and the bookings made through them — create a new booking link, change its availability rules, list confirmed bookings on it. It does not read your underlying calendar. The full install guide for Claude Code, Claude Desktop, Cursor, Windsurf, Zed, and VS Code lives in the Carly MCP post.
Carly’s MCP server today is local (stdio), so connecting it to ChatGPT takes one extra step: bridge it to a remote endpoint. Two paths:
Path A: bridge a local server with mcp-remote. mcp-remote wraps a local stdio MCP server in an HTTPS-reachable layer. You run it on a host you control (a tiny VPS, a Cloudflare Worker, your home machine over a tunnel), point it at carly mcp, and use the resulting URL as the connector endpoint. It works, but it’s a real bit of plumbing — you’re now running a server.
Path B: use a hosted Carly MCP endpoint. If a hosted endpoint is live at the time you read this — https://mcp.usecarly.com is the planned URL — you’ll be able to paste it straight into ChatGPT’s Add custom connector dialog, run through OAuth, and skip the bridge entirely. Check the Carly MCP post for current status.
If a hosted endpoint isn’t live yet, the simplest move today is to use Carly in Claude Desktop or Claude Code — see the install guide — and watch for the ChatGPT-ready hosted version.
The generic flow once you have a remote URL:
- Open ChatGPT → Settings → Connectors → Add custom connector.
- Paste the MCP server URL.
- Authenticate (OAuth flow inside ChatGPT, or paste an API token if the server uses static auth).
- Pick which tools to enable — you can scope down what the model is allowed to call.
- Save.
Now in any new chat, Carly’s tools are available. You can ask:
Make me a 30-minute “intro call” booking page with Google Meet, Mon–Thu 10–4.
Tighten my consult page to 15-minute slots only.
Who has booked through my intro call page this week?
ChatGPT will call the corresponding Carly MCP tools and return real answers, the same way Claude does today.
Best MCPs to try in ChatGPT right now
Most of the MCP servers worth using in ChatGPT today are ones that already publish hosted/remote endpoints. A starting set:
- Cloudflare — official remote MCP servers for Workers, R2, KV, and observability. Hosted by Cloudflare.
- Stripe — official Stripe MCP server for payments, customers, invoices.
- Linear — official Linear MCP server, OAuth-based, great for “what’s blocking my team this week.”
- GitHub — GitHub’s hosted MCP server covers issues, PRs, code search.
- Notion, Asana, Intercom, Sentry — all ship official remote MCP servers in 2025–2026; pull the URLs from each vendor’s docs.
For a wider catalog, see the official MCP server registry. And for the protocol itself, what is MCP? covers the architecture without the install commands.
Limitations and gotchas
A few things will trip you up:
- No local stdio servers. If a server’s setup says
command: "node"orcommand: "npx", ChatGPT can’t run it directly. Bridge it withmcp-remoteor wait for a hosted version. - OAuth happens inside ChatGPT. Servers that use OAuth will prompt you in a popup the first time. Make sure popups aren’t blocked, and re-auth if your token expires.
- Tools aren’t available in shared chats. Custom connectors are scoped to your account. If you share a conversation, the recipient won’t be able to re-run the tool calls.
- Rate limits apply. ChatGPT enforces its own per-tool rate limits on top of whatever the MCP server enforces. High-volume agentic loops will hit ceilings faster than they would in Claude Code.
- Team/Enterprise admin controls. On Team and Enterprise plans, admins can disable Developer Mode org-wide or allowlist specific connectors. If your toggle is missing, it’s likely a workspace policy.
- Apps SDK ≠ Custom Connectors. If you’re trying to publish a connector for other people to use, you want Apps SDK and its review process — pasting your URL into a friend’s ChatGPT works, but it’s not how you ship to the catalog.
For installing MCP in Claude, Cursor, or Windsurf instead, see the install guide.
Ready to automate your busywork?
Carly schedules, researches, and briefs you—so you can focus on what matters.
Get Carly Today →Or try our Free Group Scheduling Tool or Free Booking Page
