Repolode
Submit
judahpaul16

judahpaul16/gpt-home

A self-hosted voice assistant for Raspberry Pi that uses LiteLLM and LangGraph to bring ChatGPT-style conversation into a smart home setup.

Open-source alternative to:Amazon AlexaGoogle Nest Hub

View on GitHubWebsite
Stars
653
Forks
68
Open issues
0
Last commit
2 months ago

What is it?

GPT Home is a voice assistant you build yourself on a Raspberry Pi or other Linux machine, positioned as a self-hosted alternative to commercial smart speakers such as Google Nest Hub or Amazon Alexa. It routes requests through LiteLLM, so the language model behind it can be any of the 100+ providers LiteLLM supports, including OpenAI, Anthropic and Google, and the API key can be set in a .env file or through a web settings page.

The assistant handles general questions, translation, weather from OpenWeatherMap or Open-Meteo, alarms, reminders, to-do lists, and CalDAV calendar access. It also controls Spotify playback and Philips Hue lights, and keeps persistent memory in PostgreSQL with pgvector. The project runs as a set of Docker Compose services: a voice backend, a FastAPI server, a React frontend, nginx as a reverse proxy, and a Spotify Connect service.

Setup is handled by a script that installs Docker and related dependencies, and a Wi-Fi configuration script is included for Pi boards without Ethernet. Audio input and output work with ALSA or PortAudio devices, and an optional small I2C or SPI display can show status.

Key features

  • 01Connects to 100+ LLM providers through LiteLLM, with the API key set via .env or the web UI.
  • 02Handles alarms, reminders, to-do lists and CalDAV calendar events.
  • 03Provides weather lookups via OpenWeatherMap and Open-Meteo.
  • 04Controls Spotify playback and Philips Hue lights.
  • 05Stores persistent memory in PostgreSQL with pgvector.
  • 06Runs as Docker Compose services with a setup script for Raspberry Pi and other Linux systems.

Use cases

  • Building a home voice assistant on a Raspberry Pi as a privacy-conscious alternative to commercial smart speakers.
  • Adding conversational control of Spotify and Philips Hue lights to a DIY smart home.
  • Experimenting with voice assistant pipelines that combine LLMs, speech input and smart home integrations.

Quick start

  1. 1

    Run the official setup script from the README (it pipes a remote script into bash, so run it yourself)

  2. 2

    Set LITELLM_API_KEY in the .env file (or via the web interface at gpt-home.local/settings)

  3. 3

    Restart the services with the new key

    docker compose down && docker compose up -d
  4. 4

    View logs to verify the assistant is running

    docker compose logs -f

Things to know

  • Designed for Raspberry Pi and Linux systems, with audio hardware such as a speaker and microphone required.
  • Needs an API key from an LLM provider, which may incur usage costs.
  • Licensed under GPL-3.0, a copyleft license that requires derivative works to be open source.

In collections

Self-hosted apps

Topics

#voice-assistant#raspberry-pi#home-automation#llm#self-hosted#litellm#langgraph#docker#ai#automation

FAQ

What is gpt-home?
A self-hosted voice assistant for Raspberry Pi that uses LiteLLM and LangGraph to bring ChatGPT-style conversation into a smart home setup. GPT Home is a voice assistant you build yourself on a Raspberry Pi or other Linux machine, positioned as a self-hosted alternative to commercial smart speakers such as Google Nest Hub or Amazon Alexa. It routes requests through LiteLLM, so the language model behind it can be any of the 100+ providers LiteLLM supports, including OpenAI, Anthropic and Google, and the API key can be set in a .env file or through a web settings page.
Is gpt-home free to use? What license does it have?
gpt-home is open source under the GPL-3.0 license. Check the license terms before using it in commercial products.
Is gpt-home still maintained?
Updated within 6 months. The last commit to gpt-home was on Sep 13, 2026.
What can I use gpt-home for?
Building a home voice assistant on a Raspberry Pi as a privacy-conscious alternative to commercial smart speakers. Adding conversational control of Spotify and Philips Hue lights to a DIY smart home. Experimenting with voice assistant pipelines that combine LLMs, speech input and smart home integrations.
How do I install gpt-home?
Run the official setup script from the README (it pipes a remote script into bash, so run it yourself) Set LITELLM_API_KEY in the .env file (or via the web interface at gpt-home.local/settings) Restart the services with the new key: docker compose down && docker compose up -d View logs to verify the assistant is running: docker compose logs -f
Where can I get gpt-home?
The source code and installation instructions are on GitHub: https://github.com/judahpaul16/gpt-home

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects