Chatty

Chatty is a browser-based interface for chatting with local web-LLM models. It’s built for people who want to try different language models without setting up a complex environment.

Run local models in the browser

Chatty focuses on launching models via web-llm. Open the page, pick a model, and start chatting right away. The UI shows your active chats, lets you quickly start new sessions, and displays the status of the selected model.

Lower VRAM options

Some models in the list include a “(1k)” suffix. These variants reduce GPU memory requirements by about 2–3 GB, which can help if you have limited VRAM but still want to test modern LLMs.

Good for experiments and everyday prompts

Chatty works well for:

• Quick questions and explanations
• Drafting short texts
• Comparing how different LLMs respond to the same prompt
• Keeping a simple chat-style workflow with saved conversation history