- Enhanced the AI Chat Completions API to support true streaming using async generators and proper SSE headers.
- Updated endpoint paths to align with OpenAI's API versioning.
- Improved logging for request details and error handling.
- Added a new AI Models List API to return available models compatible with chat completions.
- Refactored code for better readability and maintainability, including the extraction of common functionalities.
- Introduced a VMH-specific Chat Completions API with similar features and structure.