How to Connect Twitch to ChatGPT via MCP

💡 No install? Use cloud.anythingmcp.com directly. Sign in, click Connectors → Twitch, paste your credentials, mint an MCP API key and you're done. No Docker, no git clone, no local server.

Twitch + ChatGPT

Let your AI work with Twitch: look up channels, live streams, videos, clips, categories and stream schedules, search channels, see follower counts and chat settings, and create clips or change your stream title, category and tags after your confirmation. Signed in through a Twitch developer application of your own.

Prerequisites

See the full setup instructions baked into the connector (visible in the in-app store when you select the connector). The required environment variables for this connector are:

TWITCH_CLIENT_ID, TWITCH_CLIENT_SECRET

Step 1 — Get credentials

1. Turn on two-factor authentication for your Twitch account (Settings → Security and Privacy). Twitch requires it to register applications.
2. Open dev.twitch.tv/console/apps and click Register Your Application. Enter any name and add the OAuth Redirect URL https://cloud.anythingmcp.com/api/mcp-oauth/callback. If you host AnythingMCP yourself, use https://your-anythingmcp-host/api/mcp-oauth/callback. Pick a category such as Other, choose Client Type: Confidential and click Create.
3. Click Manage on the new application. Copy the Client ID into TWITCH_CLIENT_ID, click New Secret and copy it into TWITCH_CLIENT_SECRET. The secret is shown once, and a new secret replaces the old one.
4. Install the connector, click Authorize with Provider on the connector page, sign in to Twitch and click Authorize.
5. Check the setup with twitch_get_users without arguments: it returns your own account and the user id the other tools ask for.

It is free and takes about 5 minutes.

Client ID on every call: Twitch wants the application's Client ID in a Client-Id header next to the token on every request. The connector adds it for you, so the Client ID must belong to the same application that signed you in.

What it reads: users and channels with title, category, tags and language; live streams with viewer counts; past broadcasts, highlights and uploads; clips; categories and the most watched categories right now; channel and category search; follower counts; stream schedules and chat settings. This is public data and works for any channel. The list of followers itself is shown only for your own channel or one you moderate; for other channels you get the total.

Asks before every change: twitch_create_clip clips a channel that is live right now, and twitch_update_channel changes your stream title, category, language or tags. Viewers see both at once, so the AI shows you exactly what will happen and waits for your yes. Only your own channel can be changed. A clip is created in the background: the AI checks it with twitch_get_clips after a few seconds, and Twitch returns an edit link that stays valid for 24 hours.

Switched off at install: twitch_send_chat_announcement posts a highlighted message that everyone in chat sees. To use it, add the scope moderator:manage:announcements in the connector's OAuth settings, authorize again and switch the tool on. It asks first as well.

Scopes: moderator:read:followers, clips:edit and channel:manage:broadcast. Everything else is public and needs no scope.

Ids, not names: most tools take numeric user ids, not login names. The AI looks them up with twitch_get_users or twitch_search_channels, and category ids with twitch_search_categories.

Staying signed in: access tokens last a few hours and are renewed automatically. The sign-in ends when you disconnect the application under Settings → Connections or change your Twitch password; then click Authorize with Provider again.

Common errors: 401: the sign-in expired, or the Client ID does not belong to the application that issued the token; authorize again, and update the Client ID or secret first if you changed them. 401 Missing scope: add the scope named in the message and authorize again. 400: a wrong parameter, or a title that AutoMod refused. 404: the channel has no schedule, or it is not live when you create a clip. 429: about 800 calls per minute per user; wait for the minute to reset.

The connector does not read or send chat messages or whispers, moderate, or handle subscriptions, bits, polls, predictions, channel points, raids, EventSub, stream keys or video deletion.

Step 2 — Install the adapter

curl -fsSL https://raw.githubusercontent.com/HelpCode-ai/anythingmcp/main/docker-compose.quickstart.yml -o docker-compose.yml
printf 'JWT_SECRET=%s\nENCRYPTION_KEY=%s\n' "$(openssl rand -hex 32)" "$(openssl rand -hex 32)" > .env
docker compose up -d

Step 3 — Add the connector in ChatGPT

Your server URL: in AnythingMCP, open MCP Servers → the server this connector is on, and copy its URL (https://cloud.anythingmcp.com/mcp/…). Use it wherever this guide shows YOUR_SERVER_ID.

ChatGPT adds MCP servers as plugins on its Plugins page, on the web and on paid plans. Developer mode is no longer needed. In a Business workspace only admins and owners can add one; in Enterprise and Edu an admin can grant the Create plugins with MCPs permission.

1. Go to chatgpt.com/plugins, click +, then Add custom MCP server.
2. Enter a name, e.g. Name: Twitch.
3. Under Connection, enter the Server URL https://cloud.anythingmcp.com/mcp/YOUR_SERVER_ID and keep Authentication on OAuth.
4. Select I understand and want to continue, then Create as a plugin. When ChatGPT asks, sign in to AnythingMCP and allow access.
5. Install the new plugin from your personal plugins, then type @ in a chat and pick it.

Available tools

Example prompts

• "Who is live in the Just Chatting category right now? Show the 10 biggest streams with title and viewer count."
• "When is my next scheduled stream, and what is it called?"
• "Find the most viewed clips from my channel in the last 30 days."
• "Set my stream title to Ranked grind until Diamond and the category to Valorant. Show me the change first."
• "How many followers does the channel shroud have, and is he live right now?"

FAQ

Does the free ChatGPT plan work with custom MCP servers? OpenAI documents custom MCP servers for paid plans only; the Free plan is not on its list.

Next steps

• Connector landing page
• Connect to Claude
• Connect to Cursor
• Connect to Gemini
• Connect to Meta Muse