minimal-chat

Introduction to MinimalChat: A Simple and Customizable LLM Chat Client

MinimalChat is an innovative chat application that is open-source and designed to be both lightweight and versatile. Its primary objective is to provide users with a robust chat experience by supporting multiple language models, including the well-known GPT-4 Omni, along with various custom and local model endpoints. The application is meticulously crafted to be simple yet fully featured, offering a responsive experience whether you are accessing it on a desktop or using it via a Progressive Web App (PWA) on your mobile device.

Self-Hosting with Docker

MinimalChat can be self-hosted with the Docker platform, allowing users to run their own instance effortlessly.

docker pull tannermiddleton/minimal-chat:latest

Try It Out

Curious about MinimalChat and how it operates? You can view the application in action by watching a demo video or a higher-quality version on YouTube.

Robust Features

MinimalChat is packed with features that make conversation intuitive and engaging. Here are some of the core offerings:

• A clean, minimalistic design for ease of use.
• Ability to interact with voiced conversations using Speech-to-Text (STT) and Text-to-Speech (TTS).
• Support for numerous language models, including custom and local endpoints using WebLLM.
• Flexibility to swap models during conversations seamlessly.
• Interactive swipe gestures for easy access to settings and previous conversations.
• Tools to edit, regenerate, or remove past messages.
• Built-in markdown support and code syntax highlighting.
• Integration with DALL-E 3 for generating images.
• Options for importing and exporting conversations.
• Adaptable, mobile-friendly layout with PWA support.

Operating Offline

MinimalChat also supports offline operation, which allows users to host a local language model through LM Studio or load a full model directly in their browser.

Secure and Private

User privacy and data security are priorities in MinimalChat. All conversations are stored safely and locally on your device, ensuring that sensitive information remains private.

Ease of Setup

Setting up MinimalChat is straightforward. Follow these steps:

1. Run npm install to install necessary packages.
2. Use npm run build to compile the application.
3. Start the local server using npm run preview for production or npm run dev for development purposes.

For detailed configuration guidance, the Wiki is an excellent resource.

Frequently Asked Questions

• Is it free? Yes, MinimalChat is completely free to use under the MIT License. However, certain API interactions may require keys.
• Can MinimalChat work offline? Absolutely. Simply use WebLLM to run your model locally.
• Mobile Compatibility? It is completely mobile-compatible and can be installed as a PWA.
• Privacy Assurance? Yes, all conversations remain locally stored on your device.

Contribution Invitation

The developers openly invite contributors to enhance MinimalChat further. Interested parties can submit issues, fork the repository, and make pull requests while adhering to established coding standards. Contributions help in refining the application, so a thorough description for any pull requests is appreciated.

Troubleshooting

Should you encounter any issues, some basic troubleshooting steps include:

• Verifying internet connectivity and API key validity.
• Clearing the browser cache as a last resort. For further assistance, issues can be reported through the issue tracker.

With its wide array of features and open-source nature, MinimalChat stands as an accessible chat client for anyone interested in language model integrations.