Can I use LM Studio, vLLM, DeepSeek, Mistral, Together or another OpenAI-compatible server with BookyAI?

Yes, from version 1.9.16. Open Settings, "AI engines", click "OpenAI", and fill in "Base URL" with the address of the server, for example http://localhost:1234/v1 for LM Studio or https://api.deepseek.com for DeepSeek. Leave it empty to use OpenAI itself. Then save the key for that server (a server on your own computer or network usually needs no key), and choose a model: the list loads from that address, and you can also type the model name by hand. Click Test. If Test says there is no chat endpoint at the address, check the Base URL: most servers end in /v1. BookyAI works with any server that speaks the OpenAI chat API, but it cannot support each third-party server individually, and the quality of the book is that of the model you run: small local models write noticeably weaker chapters.