GPT4o-mini as default

[briansunter]

As of July 2024, gpt-4o-mini should be used in place of gpt-3.5-turbo, as it is cheaper, more capable, multimodal, and just as fast. gpt-3.5-turbo is still available for use in the API.

https://github.com/briansunter/logseq-plugin-gpt3-openai/issues/new

[Mieluoxxx]

Hey bro, I’ve been checking out DeepSeek—it’s crazy cheap and plays nice with the OpenAI API. Wanna give it a shot?

[briansunter]

Deepseek is great. It should be compatible if you run it locally with something like ollama and update the plugin to use the local url. For the model it uses by default, I'll probably stick with the openai models for now though