June 5, 2025
I couldn't remember where my talk at @aiDotEngineer is this morning, so I asked the voice agent on the website.
We generally don't recommend using preview models in production. But for AI Engineer, we figured that using the newest Gemini 2.5 Flash native audio model was the right choice. Try it out and see what you think.
And if you're interested in voice AI, I'll be talking about deploying open source @pipecat_ai agents to production at enterprise scale. (At 11:00am)