HuggingChat capabilities
9 mapped capabilities, each graded and dated. The map shows what HuggingChat can do; the audit shows whether it’s worth consolidating — and a guide shows how to move.
Capabilities
Custom system prompt and conversation settings
provisionalverified ~2 months agoUsers can set a custom system prompt to define the assistant's role, tone, and formatting, plus adjust per-conversation behavior.
Guest access plus signed-in chat history
provisionalverified ~2 months agoHuggingChat can be used without an account as a guest; signing in with a Hugging Face account saves chat history and personal settings.
Hugging Face Hub MCP server (search models, datasets, Spaces, papers)
provisionalverified ~2 months agoAn official Hugging Face MCP server that connects a chat assistant to the Hugging Face Hub so it can search and explore models, datasets, Spaces, and papers, and run community Gradio tools.
MCP tool calling (agentic tools)
provisionalverified ~2 months agoLets the model call external tools exposed by Model Context Protocol (MCP) servers during a conversation and feed the results back into its answer.
Multimodal image input (vision)
provisionalverified ~2 months agoUsers can attach images to a message so the model can analyze or answer questions about them.
Omni automatic model router
provisionalverified ~2 months agoAn automatic router (branded 'Omni') that selects an appropriate underlying model for each message instead of requiring the user to pick one manually.
Open-source LLM catalog with manual model selection
provisionalverified ~2 months agoA free chat assistant that lets users pick from a large catalog of open-source / open-weight large language models and switch between them per conversation.
Open-source, self-hostable chat app (chat-ui)
provisionalverified ~2 months agoThe application powering HuggingChat is open source and can be self-hosted, configured against your own OpenAI-compatible model endpoints.