Claude Chat for Home Assistant (Marketplace integration)

Installs: 52 sharing analytics
GitHub stars: 5
Connection: Uses the vendor's cloud
Upkeep: Updated in the last 6 months
Version: v0.7.0, September 7, 2026
Category: AI and voice, AI models
Kind: Marketplace integration
Brand: Anthropic
Needs: Needs the Marketplace (HACS on 2026.10 and older)
Domain: claude_chat
Source: github.com/h0jeZvgoxFepBQ2C/claude_chat_for_homeassistant
Issues: Report an issue

Independent directory, made by the developer of Vulpo. Not affiliated with Home Assistant or the Open Home Foundation.

Screenshots From the README of h0jeZvgoxFepBQ2C/claude_chat_for_homeassistant.

About

Claude Chat adds a sidebar chat panel to Home Assistant that is powered by Anthropic's Claude. It lets the model inspect entities, areas, dashboards and automations, and read automation traces and state history to explain why an automation did not fire. Changes to dashboards, service calls and automations are staged first, shown as a diff, a payload or YAML, and applied only after approval. Chats are kept as multiple persistent sessions with streamed responses, a model choice per session, and support for dropping in images. It relies on Anthropic's cloud and is set up in the interface with an API key and a model choice. The README marks it as alpha, and YAML-mode dashboards stay read-only.

Automatic summary of the README and code of h0jeZvgoxFepBQ2C/claude_chat_for_homeassistant. Something wrong, or are you the maintainer and want your own text here? Get in touch

Supports

Uses the vendor's cloud Needs an internet connection and an account with the vendor.
From the Marketplace Installed through the Marketplace, formerly HACS.
Actively maintained Its author released or pushed a change in the last six months.

Setting it up

Install it from the Marketplace and restart Home Assistant, then add it under Settings, Devices and services.

Claude Chat

API key
Model

After setup

Model

Texts from the code of h0jeZvgoxFepBQ2C/claude_chat_for_homeassistant.

From the README

⚠ Alpha — usable, feedback welcome. This is a new integration; please run it on a non-critical instance for now and open issues for anything unexpected. End-to-end flows are covered by tests and live verification, but it hasn't yet had time to soak in the community.
Unofficial community integration. Not affiliated with or endorsed by Anthropic. "Claude" is a trademark of Anthropic, used here descriptively to identify the supported model.
Adds a Claude Chat entry to the HA sidebar.
Multiple persistent chat sessions, auto-titled after the first message (like ChatGPT).
Streaming responses over HA's websocket — markdown, code blocks, and tables render live as Claude types.
Model picker per session (Haiku / Sonnet / Opus); the choice is remembered.
Stop button to cancel a streaming response mid-flight.
Claude has a small, scoped set of HA tools:
list_entities, get_entity, list_areas
list_dashboards, get_dashboard, list_lovelace_resources
list_automations, get_automation, list_services
Debugging: list_automation_traces, get_automation_trace, get_state_history — ask "why didn't my X automation fire yesterday" and Claude pulls the trace, reads the conditions/triggers, and explains
propose_dashboard_update, propose_service_call, propose_automation_create, propose_automation_update, propose_automation_delete — all stage changes for your approval
Dashboard edits show a unified diff. Service calls show the payload. Automations show the YAML. Nothing is applied without you clicking Apply.