The team at @AssemblyAI released a new realtime transcription (speech-to-text)…

June 2, 2025

The team at @AssemblyAI released a new realtime transcription (speech-to-text) model today. It's fast and accurate. Very much worth testing if you are building voice AI applications.

Shoutout to the engineers at AssemblyAI for the excellent work on this model, the APIs and infrastructure that serve it, and the new @pipecat_ai `AssemblyAISTTService` implementation. 🎶

We're co-hosting a Voice Agents happy hour on Wednesday evening. Whether you're in town for @aiDotEngineer World's Fair, or you live in San Francisco, please join us. Looking forward to seeing new friends and old.

RSVP link in 🧵

Wednesday Voice Agents Happy Hour link^[1]

@AssemblyAI new realtime text-to-speech API launch thread ...

https://t.co/obVYXrDi9d

AssemblyAI@AssemblyAI

Introducing Universal Streaming - an ultra-fast, ultra-accurate streaming speech-to-text model for voice agents 🚀

Universal-Streaming delivers ultra-low latency, superior accuracy, and intelligent endpointing at just $0.15/hr 👇