An AI browser assistant addon for Firefox that can read the text on the screen and have a chat about it. Tested with LLM backend Ollama assumed to be hosted at http://localhost:11434.
- Talk to Ollama
- Send text on screen as part of input context
- Use vision model to dynamically generate enriched alt text for images to be sent as text to the input context.
- Make model and endpoint configurable
- Use langchain to chunk and embed the page for real-time RAG
Load the Extension in Firefox:
- Install Firefox Developer Edition
- Open Firefox and navigate to
about:debugging - Click on "This Firefox" on the left hand side and then "Load Temporary Add-on".
- Select the
manifest.jsonfile from your extension directory.
Be sure to set OLLAMA_ORIGINS=* environment variable. See "how do I configure Ollama server?"