[Feature] Response streaming

[imaginenotusingthis]

Is it possible to add streaming to the AI response? Sometimes it takes forever to get a response and I end up hoping it finishes soon. Streaming would at least help with bordem.

[Draculabo]

Hello, could you elaborate on your usage scenario? Where are you planning to use the streaming response? Do you need support for SSE or WebSocket? Do you have requirements for integration with Codex? I am currently working on adapting to Codex.