This project provides an open-source, SvelteKit-based chat interface designed to power applications like HuggingChat. It enables users to interact with various open-source and commercial Large Language Models (LLMs), offering features like web search integration and customizable model configurations.

How It Works

The UI connects to LLMs via API endpoints, supporting numerous providers including OpenAI-compatible services, Anthropic, Google Vertex AI, Cloudflare, and local inference servers like llama.cpp and Ollama. It leverages a MongoDB instance for storing chat history. A key feature is its integrated web search capability, which uses text embeddings to retrieve relevant information for Retrieval-Augmented Generation (RAG).

Quick Start & Requirements

• Docker Deploy: docker run -p 3000 -e HF_TOKEN=hf_*** -v db:/data ghcr.io/huggingface/chat-ui-db:latest
• Local Setup: Requires Node.js, MongoDB, and optionally llama.cpp for local LLM execution.
• Docs: hf.co/docs/chat-ui

Highlighted Details

• Supports a wide array of LLM providers and local inference backends (TGI, llama.cpp, Ollama).
• Integrated web search with customizable embedding models (transformers.js, TEI, OpenAI).
• Extensible configuration for models, prompts, and parameters via .env.local.
• Supports multimodal models and advanced features like OpenID Connect authentication.

Maintenance & Community

The project is actively maintained by Hugging Face. Community support channels are not explicitly mentioned in the README.

Licensing & Compatibility

The project is licensed under the AGPL-3.0 license. This is a strong copyleft license, requiring derivative works to also be open-sourced under AGPL-3.0.

Limitations & Caveats

The AGPL-3.0 license may impose restrictions on commercial use or linking with closed-source applications. Running HuggingChat locally requires significant setup, including a Hugging Face Pro account for some models and API keys for services like Serper.dev.