Local embedding models / EmbeddingGemma

[krzemian]

With EmbeddingGemma and Qwen models looking promising, I'd be heavily interested in doing embeddings locally. Currently it's only possible via API.

Does Obsidian Copilot plan to allow local models for this?

I'm also open to having my stance challenged as to why it would not be that risky to sent all my notes through to a 3rd party.

[logancyang]

You can use LM Studio or Ollama with copilot. Please check our docs for instructions.