DOCSEARCHbyAlgolia
Submit Docs

Typing Mind Documentation

repository·main·Indexed 21 days ago

https://github.com/typingmind/typingmind

An advanced, offline-first Chat UI frontend for interacting with AI models such as ChatGPT, Claude, and Gemini using personal API keys. This documentation covers self-hosting the static web application, connecting to custom endpoints or local models (e.g., Ollama, LocalAI), and managing license and API key requirements.

Tokens
1.1K
Snippets
4
Records
6
Agent score
26%

What's inside Typing Mind

  1. Understand Self-Hosting for TypingMind

    main

    Self-hosting involves deploying the compiled version of TypingMind on your own web hosting server or domain for personal use. This provides privacy and availability, as you can access your own version even if the main TypingMind.com service is inaccessible.

    Key constraints and details:

    • Deployment: You can deploy to any server or static web cloud service (e.g., GitHub Pages, Cloudflare Pages, AWS S3, Vercel, Netlify, Heroku) or run it on localhost.
    • Customization: The self-host version is identical to the version on TypingMind.com and cannot be customized. For custom branding or team management, use TypingMind Custom.
    • Source Code: You receive the compiled code, not the full source code. The app is not open-source. You have permission to deploy and use the compiled code, but you do not have permission to modify, redistribute, share, or resell it.
    • Requirements: You still require both a TypingMind License Key (to verify via TypingMind's license server) and an OpenAI API Key to use the application.

  2. Connect to local or custom AI models

    main

    You can extend Typing Mind to work with custom endpoints, third-party providers (like OpenRouter), or locally hosted models (like Ollama or LocalAI) by following these steps:

    1. Open the Typing Mind application.
    2. Click the model dropdown in the UI.
    3. Navigate to Model Settings $\rightarrow$ Add Custom Models.
    4. Enter the required details:
      • Endpoint: The URL of your model provider.
      • Model ID: The specific identifier for the model.
      • Note: Custom headers and body parameters are also supported.
    5. Click Test to verify the connection.
    6. Click Add Model to save it.

    Troubleshooting CORS: If you encounter CORS (Cross-Origin Resource Sharing) issues when connecting to a local model, ensure that your local model server is configured to run in CORS mode.

    Supported integration types mentioned in documentation include:

    • Local AI (LLama, GPT4All, etc.)
    • Azure OpenAI
    • OpenRouter
    • OpenAI Fine-tuned models
    • Mixtral AI
    • Perplexity AI
    1. Open the app
2. Click the model dropdown → Model Settings → Add Custom Models
3. Enter your custom model details: Endpoint, Model ID.
4. Click "Test" and then "Add Model".

  3. Update Typing Mind self-hosted version

    main

    To update your local installation to the latest available version in the repository, perform a git pull and restart your server:

    git pull
# Then restart your server (e.g., yarn start)

    Note: The static self-host version updates are typically delayed by 1 or 2 versions compared to the latest versions available at www.typingmind.com.

    git pull

  4. Run Typing Mind locally

    main

    To run the static self-host version of Typing Mind on your local device or a private server, follow these steps:

    1. Clone the repository.
    2. Install dependencies using yarn or npm.
    3. Start the development server.
    4. Access the application at http://localhost:3000.

    Important Security Note: If you host the application on any hostname other than localhost, you must use HTTPS to ensure all application features function correctly.

    Alternatively, since this is a static version, you can use any static web server to host the src folder directly.

    ```bash
# Clone the repo
git clone <repository-url>

# Install dependencies
yarn install
# or
npm install

# Start the server
yarn start
# or
npm run start
```埋入```

  5. Update a Self-Hosted TypingMind instance

    main

    To update your self-hosted version to the latest release, clone or pull the latest changes from the official GitHub repository and then restart your server.

    git pull origin main
# Restart your server after pulling changes

  6. Requirements for Self-Hosted TypingMind

    main

    To run a self-hosted instance of TypingMind, you must provide the following credentials within the application:

    1. TypingMind License Key: Used to verify your license via the TypingMind license server.
    2. OpenAI API Key: Required to power the AI models.

    Note on Security: Do not attempt to embed your License Key or API Key into the self-hosted version to share with others. The self-host version is intended for personal use. For team or community deployments, use TypingMind Custom.