chat-ui and svelte-chat-ui
These are ecosystem siblings—one is the production chat interface powering HuggingFace's service while the other is a standalone SvelteKit implementation that developers can self-host or integrate into their own applications, both built on similar web frameworks but serving different deployment contexts.
About chat-ui
huggingface/chat-ui
The open source codebase powering HuggingChat
A SvelteKit frontend that communicates exclusively with OpenAI-compatible APIs, enabling unified access across providers (Ollama, llama.cpp, OpenRouter, etc.) via a single configuration. Stores chat history, user data, and settings in MongoDB, with optional smart routing via an embedded Arch routing model to dynamically select the best LLM per request. Supports Model Context Protocol (MCP) servers for tool integration and multimodal requests, alongside pluggable theme customization and deployed either locally or via Docker.
About svelte-chat-ui
iansinnott/svelte-chat-ui
Chat with LLMs. Bring your own key. Keyboard-centric, mobile friendly, searchable.
Supports full-text search across chat history with local SQLite storage via WASM-compiled CRDTs for conflict-free cross-device sync. Available as both a web app and Tauri-based desktop client, enabling use with OpenAI, local LLMs (Llama, Mistral), and custom system prompts. Built with Svelte for reactive UI and includes annotation features to tag responses with notes.
Related comparisons
Scores updated daily from GitHub, PyPI, and npm data. How scores work