Integration with other LLM models besides OpenAi

[anibalbezerra]

Dear developers,

First of all, a huge shout-out for such an amazing project! I've read the docs and was unable to discern if the Memori module works with other LLM providers. I'm not referring to the integration (watching side) with the LLM in the project running Memori; I'm talking about the memory management per se.
Thanks for your time.

Best

Anibal

[harshalmore31]

Hi @anibalbezerra! Thanks for the appreciation !
Yes, Memori works with multiple LLM providers! You can use OpenAI, Azure OpenAI, Ollama (local models), and 100+ other providers via LiteLLM integration. It also supports any OpenAI-compatible endpoint, which covers most modern LLM providers. However, Anthropic (Claude) and some direct provider SDKs aren't supported yet.

The setup is straightforward - just configure your endpoint URL and API key, then call memori.enable() for universal recording. Your application can use any supported LLM provider for conversations, while Memori's internal memory processing agents use OpenAI or given OpenAI-compatible provider for analyzing and categorizing the conversations.

Hope this clarifies the integration! Let me know if you need help with a specific provider setup.