❤️ how OpenAI is doing several things at once in the release today:

March 20, 2025

❤️ how OpenAI is doing several things at once in the release today:
➡️ advancing the state of the art in native audio capabilities
➡️ giving voice AI developers multiple, modular ways to use their models/features, and
➡️ building first-class support for voice into their agent SDK, including traces and dashboard tooling.

Peter Bakkum@pbbakkum

I'm very excited that the popular Pipecat project has refreshed their integrations with the OpenAI APIs, it's a great way to access Realtime speech2speech, speech-to-text, and text-to-speech