still working on the response rendering, but was able to get dave working with my wireguard ollama instance (on a plane!)
private, local, nostr ai assistant.
👀
is this its own ollama model or can you run it on smth small like gemma3:1b?
Please Login to reply.
you can run it on any openai-compatible AI backend with tools support
in this case I was using some random tools-supported model I had on my ollama server:
hhao/qwen2.5-coder-tools:latest
oh cool!
