> For the complete documentation index, see [llms.txt](https://docs.podigee.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.podigee.com/documentation/start-here/ai-agents/start-here/choose-how-to-use-podigee-for-ai-agents.md). # Choose how to use Podigee for AI agents Choose the Podigee AI connection that matches your client, media, and automation needs. Choose a compatible Model Context Protocol (MCP) client. Claude, ChatGPT, Codex, Claude Code, Cursor, VS Code, Gemini CLI, and other MCP clients connect with browser sign-in. Connecting AI assistants is part of each current Podigee plan, starting with Launch. Podigee is listed in the Claude Connectors directory and in the ChatGPT plugin directory. In Podigee, **Data & integrations > Connected apps** shows the fastest way to connect each common assistant. ## Choose a connection | Your work | Recommended connection | Why | | ------------------------------------------------- | ---------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------- | | Use Podigee in ChatGPT | The Podigee plugin from the ChatGPT plugin directory | You install the plugin and approve the account in the browser. Developer mode is not necessary. | | Use Podigee in Claude | The Podigee connector from the Claude Connectors directory | You connect the listed connector and approve the account in the browser. It also works in Claude Code. | | Work in a terminal with Codex | Codex 0.154.0 or later with browser sign-in | Codex runs on your computer and receives the Podigee authorization on a local address. An earlier version cannot sign in. | | Work in a terminal with Claude Code or Gemini CLI | Browser sign-in from the terminal | The tool opens the Podigee sign-in in your browser and receives the authorization on a local address. | | Work in Cursor or VS Code | **Add to Cursor** or **Add to VS Code** in Podigee | The link opens the editor with the Podigee address filled in. You approve the account in the browser. | | Publish a local audio or video file | An assistant that runs on your computer, for example Claude Code, Codex, or Cursor | The assistant can read the local file and send it directly to Podigee. Nothing is installed. | | Use another AI application | Compatible remote MCP client | The client can discover the Podigee MCP server and request access through OAuth. | | Run an unattended script that calls MCP tools | MCP client with a personal access token | A token avoids an interactive browser login. Store it as a secret. | | Upload media from your own script | Podigee REST API with an API key | The REST API has its own upload endpoints. See the API Reference. | You can use more than one connection. Each authorization applies to one selected Podigee account or organization. ## Use the directory listings Podigee publishes its own listings: * Claude: [claude.ai/directory/podigee](https://claude.ai/directory/podigee). See [Connect Podigee to Claude](/documentation/start-here/ai-agents/connect-and-authorize/connect-podigee-to-claude.md). * ChatGPT and Codex: the [Podigee plugin](https://chatgpt.com/plugins/plugin_asdk_app_6ac38ec038088191b1134a7909638800). See [Connect Podigee to ChatGPT](/documentation/start-here/ai-agents/connect-and-authorize/connect-podigee-to-chatgpt.md). Each listing names Podigee GmbH as the developer and uses the MCP address `https://mcp.podigee.com/mcp`. Do not connect a listing or a custom connector with the Podigee name that shows a different developer or a different address. A listing removes the manual setup. You can still add the Podigee MCP address by hand in a client that has no directory, or when your workspace does not show the directory. ## Use an assistant on your computer for local media Use an assistant that can read files and run commands when the workflow starts with a file on your computer. The assistant asks Podigee for an upload address, sends the file with one HTTPS `PUT`, and continues with the same protected Podigee workflows as other clients. You do not install a Podigee program for this. See [Upload local media with your assistant](/documentation/start-here/ai-agents/transfer-media-encode-and-publish/upload-local-media-with-your-assistant.md). ## Use another MCP client for portability Podigee exposes a remote MCP server at `https://mcp.podigee.com/mcp`. A compatible client can discover the authorization flow, ask you to approve access, and use the Podigee capabilities that match the granted scopes. Podigee accepts each MCP client that supports remote Streamable HTTP servers and OAuth. It does not keep a list of permitted clients. The Podigee authorization page shows which app requests access before you approve it. ## Local media boundary Chat messages and MCP tool calls do not carry large media files. A hosted connector cannot read a file from your computer unless the application provides a separate local execution environment. In a chat assistant, upload the file in Podigee on the **New episode** page. You can then continue the episode workflow from any authorized client. Podigee does not fetch episode media from a remote URL through MCP. ## Next step [Connect and run your first read-only check](/documentation/start-here/ai-agents/start-here/connect-and-run-your-first-read-only-check.md) before you grant write access or prepare a publication. --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation by asking a question. Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters: ``` GET https://docs.podigee.com/documentation/start-here/ai-agents/start-here/choose-how-to-use-podigee-for-ai-agents.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.