Home
Use this wiki to install the Groq integration, add named AI services, choose the right Home Assistant surface, and troubleshoot Groq API or model issues.
Important
This is an unofficial community project. It is not affiliated with, endorsed by, or supported by Groq.
Groq model availability, account limits, rate limits, and feature support can vary by Groq account, project, and model.
| Task | Start here |
|---|---|
| Install through HACS or manually | Installation |
| Add an API key and named service entries | Account and Service Setup |
| Use Assist, AI Tasks, Home Assistant tools, response actions, structured outputs, reasoning, prompt caching, or Compound built-in tools | Text Generation |
| Use Groq in a Home Assistant voice pipeline | Speech-to-Text |
Generate speech through tts.speak, including Groq output formats, sample rate, speed, and Long TTS |
Text-to-Speech |
| Analyze camera snapshots, media images, local images, image URLs, OCR prompts, or Qwen vision models | Image Recognition |
| Fix auth, model, rate-limit, TTS, or diagnostics issues | Troubleshooting |
| Run tests or contribute changes | Development |
- Home Assistant
2026.6.0or newer. - A Groq API key from the Groq Console.
- Network access from Home Assistant to
https://api.groq.com. - Optional:
ffmpegon the Home Assistant host if you enable TTS audio normalization, Long TTS, or processed playback conversion.
Groq response actions require explicit targets. Use service_id for service-level actions such as groq.generate_text, groq.analyze_image, and groq.transcribe_audio. Use config_entry_id for account-level actions such as groq.clear_cache and groq.list_models.