Chrome browser extension for chatting about private pages with local LLMs (github.com)

A new Chrome browser extension has been unveiled, allowing users to chat with Local Language Models (LLMs) powered by OpenAI-compatible APIs directly from a sidebar interface. The extension, currently in its 0.1-beta version, enables users to include the content of the currently loaded web page by converting its HTML to Markdown, thus enhancing the contextual relevance of their queries. This tool is particularly significant as it addresses previous limitations faced by users when attempting to interact with page content in real-time, making it easier to extract and analyze specific information from dynamic pages. Designed with user-friendly features like real-time streaming responses, persistent settings, and a straightforward configuration process, the extension is built to operate without remote dependencies. It requires a robust model with at least a 16k context, ideally 32k, to ensure effective handling of larger web pages. The simplicity of the installation process and the local execution of LLMs through platforms like Ollama further emphasize its accessibility, providing a valuable resource for researchers, analysts, and everyday users interested in leveraging AI for web interactions.