Does the new WebRTC requires (for Realtime voice) VAD implement notation or is that handled by OpenAI backend?
For finetuning grading, could there be an option for emitting an HTTP request to us and receiving the evaluation back?
We plan to improve the Assistants API next year. You can continue to use our code_interpreter tool in your apps and we’ll continue to ensure its performance!
why switch from system prompt to developer message for o1? not even a soft backward compatibility when its the same structure via api reference?
best. mini dev day. ever.
happy x-mas api team!
Thanks for tuning in (and using OpenAI)!
We definitely plan on supporting o1 in Assistants. We were waiting to add support for tools before we supported it in Assistants. Now that function calling is enabled in o1, o1 is coming to Assistants soon!
You can inspect the text returned by the vector store today! see Inspecting file search chunks here https://platform.openai.com/docs/assistants/tools/file-search#improve-file-search-result-relevance-with-chunk-ranking. We are also working on more ways to work with vector stores - more to come early next year.
We are working on all of this 
What kind of microcontroller is compatible with the Realtime API?
I understand this has taken longer than expected but we plan to iterate on and improve the Assistants API next year. In the meantime, we’re committed to making sure that your apps on the Assistants API work performantly.
Are there plans to allow fine-tuning the multi-modal models?
If there are plans, what would the fine-tuning process look for something like audio or images? (BASE64 images / audio in JSONL?)
Sounds like a cool idea – we’ll discuss it with the team!
Not immediately, but this is an interesting idea!
Yes, it is available in Assistants
- Vector stores can’t be used outside the Assistants API at the moment, but we’ll take this as a feature request!
- No limits on number of Assistants.
This is a really useful feature and we’ll work on bringing this to the OpenAI API.
A high reasoning_effort tells the model to think longer in a single model turn.
o1-pro uses techniques that go beyond thinking for longer and we’re planning on supporting o1 pro mode in the API!
VAD functionality is automatically handled when using the WebRTC API.
o1-pro uses techniques that go beyond thinking for longer. We’re working on bringing o1 pro mode to the API soon!