|
Hi,
Today we're launching Cekura Bench: open benchmarks for voice AI models.
We built this for you, and for everyone building in voice. "Which model should we use?" is one of the
questions we hear most often from customers, so we tested the leading speech-to-speech and cascaded
models side by side on real conversational tasks: turn-taking, interruptions, backchannels and task
completion.
Any support from your side would go a long way 🙏
Thank you for building with us.
Best, Tarush & the Cekura team
|