Privacy & Support
Privacy details for each LLM backend and where to report issues.
Privacy
Selora AI Cloud: When using Selora AI Cloud, device names, entity names, entity states, and automation text are sent to Selora Homes’ managed LLM gateway for processing. No audio or video is ever transmitted. Selora Homes does not retain prompts or completions for training, and access is scoped to your authenticated Selora Homes account. Review the Selora Homes privacy policy for full details on data handling and retention.
Anthropic Claude: When using the Claude backend, device names, entity names, entity states, and automation text are sent to the Anthropic API for processing. This includes the names of your automations and rooms. No audio or video is ever transmitted. Review Anthropic’s privacy policy for full details on data handling and retention.
OpenAI: When using the OpenAI backend, the same data (device names, entity names, entity states, and automation text) is sent to the OpenAI API for processing. No audio or video is ever transmitted. Review OpenAI’s privacy policy for full details on data handling and retention.
OpenRouter: When using the OpenRouter backend, the same data is forwarded to OpenRouter and then to the upstream model provider you have selected. Review OpenRouter’s privacy policy and the policy of the upstream provider for full details on data handling and retention.
Google Gemini: When using the Google Gemini backend, the same data is sent to the Google AI API for processing. No audio or video is ever transmitted. Review Google’s privacy policy for full details on data handling and retention.
Ollama: All data stays on your local network. Nothing is sent outside your home.
Issues & Support
Report bugs or request features on the GitHub issue tracker .
Last modified May 6, 2026: Document Selora AI Cloud and additional LLM providers (41a6d34)