Voxel
An open-source desktop voice changer that feeds any app through a virtual mic
Emerged finds new products worth a look, daily — what’s new today
Voice agents mostly feel wrong because they interrupt or hesitate, and this treats that as a conversation problem rather than a noise-cancellation one.
Sparrow-2 is a turn-taking model from Tavus for real-time conversational AI. Rather than filtering the audio, it reads the conversation: whether the person has finished a thought, whether an interruption is a genuine bid to speak, whether a sound is a backchannel like "mm-hmm", and whether noise in the room is someone else entirely. Tavus reports it as the top-ranked model on TurnBench. There is a live demo on the site, with documentation, the underlying research and access through the Tavus developer platform. Pricing is not published on the page.
More like this
Nico
Hold a key, say what you need, and a sandboxed worker on your Mac does it
Lokutor
Voice agents that run entirely on CPUs, in the cloud, on-prem or on-device
Egma
Open-source simulation testing and monitoring for voice agents
JARVIS
A voice assistant that brainstorms with you, then runs and watches Claude Code
Emma
Video-call an AI character that listens and talks back at once, in your browser