LLM Assistant

About

Unlock the full potential of Large Language Models directly in your browser. Universal LLM Assistant is a privacy-first, highly customizable AI chat extension designed for developers, researchers, and power users who want full control over their AI tools.

Instead of being locked into a single provider, this extension allows you to "Bring Your Own Key" (BYOK) and connect to any OpenAI-compatible API endpoint (including DeepSeek, Groq, local models via LM Studio or Ollama, and more).

Key Features:

Custom API Endpoints: Easily configure your own API URL, API Key, and Model Name in the sidebar. Switch between different models on the fly.

Persistent Local Sessions: Your chat history is saved entirely on your local machine using Chrome's local storage. Revisit past conversations seamlessly without relying on third-party servers to store your data.

Vision & File Attachments: Upload images (processed via Vision API), code files (.py, .js, .json), and text documents (.csv, .md) directly into the chat context.

Real-Time Streaming: Enjoy smooth, token-by-token streaming responses with a smart "Scroll Lock" feature that lets you read previous messages without being forcefully yanked to the bottom.

Granular Control: Fine-tune the AI's creativity using the built-in Temperature slider (0.0 to 2.0).

Prompt Rollback: Made a mistake or want to branch a conversation? Hover over any past prompt to instantly truncate the chat history and roll back to that exact moment.

True Incognito Support: Fully supports Chrome's Split Incognito mode. Open the chat in a private window and all traces of that specific conversation will vanish when the window is closed.