Extraction model

Connect with your API key to bring your own extraction model.

Extraction model

Bring your own model. RememberOS extracts facts and relationships from your memories using a language model. Point it at your own OpenAI-compatible endpoint — your OpenAI/Azure key, or a self-hosted Ollama / llama.cpp — for full cost control and data residency. The key is connection-tested with a one-token completion, encrypted at rest, and never returned. Leave it on Platform to use ours.

Extraction provider

Any OpenAI-compatible chat-completions endpoint. Your memory text is sent only to this endpoint for extraction.

Platform

Examples: OpenAI https://api.openai.com/v1 + gpt-4o-mini · Ollama http://localhost:11434/v1 + llama3.2 · Azure / Mistral / DeepSeek / any OpenAI-compatible gateway.