Skip to content

Fully private LLM chatbot that runs entirely with a browser with no server needed. Supports Mistral and LLama 3.

License

Notifications You must be signed in to change notification settings

yuichiromukaiyama/secret-llama

 
 

Repository files navigation

Secret Llama

secret llama

Entirely-in-browser, fully private LLM chatbot supporting Llama 3, Mistral and other open source models.

  • Fully private = No conversation data ever leaves your computer
  • Runs in the browser = No server needed and no install needed!
  • Works offline
  • Easy-to-use interface on par with ChatGPT, but for open source LLMs

Big thanks to the inference engine provided by webllm.

Join us on Discord

System Requirements

To run this, you need a modern browser with support for WebGPU. According to caniuse, WebGPU is supported on:

  • Google Chrome
  • Microsoft Edge

It's also available in Firefox, but it needs to be enabled manually through the dom.webgpu.enabled flag. Safari on MacOS also has experimental support for WebGPU which can be enabled through the WebGPU experimental feature.

In addition to WebGPU support, various models might have specific RAM requirements.

Try it out

You can try it here.

To compile the React code yourself, download the repo and then, run

yarn
yarn build-and-preview

If you're looking to make changes, run the development environment with live reload:

yarn
yarn dev

Supported models

Model Model Size
TinyLlama-1.1B-Chat-v0.4-q4f32_1-1k 600MB
Llama-3-8B-Instruct-q4f16_1 ⭐ 4.3GB
Phi1.5-q4f16_1-1k 1.2GB
Mistral-7B-Instruct-v0.2-q4f16_1 ⭐ 4GB

Looking for contributors

We would love contributions to improve the interface, support more models, speed up initial model loading time and fix bugs.

Other Projects by Author

Check out screenshot to code and Pico - AI-powered app builder

About

Fully private LLM chatbot that runs entirely with a browser with no server needed. Supports Mistral and LLama 3.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • TypeScript 85.6%
  • JavaScript 8.3%
  • CSS 5.0%
  • HTML 1.1%