Search
Have an existing account?
Sign In
Tag: speech models
Block built a voice model that never touches text
Dessa runs on a single GPU, learns only from recorded conversation, and carries emotion in the audio rather…
Alibaba’s Qwen adds a full-duplex voice stack for agents
The realtime model decides whether to listen, speak or wait, and ships only through a managed API.
ElevenLabs rebuilds its voice stack around a faster v4 model
The new speech family adds directional control, longer productions and a 150 millisecond voice agent tier.
Kyutai teaches a speech model to do arithmetic out loud
Kyutai released two open-weight speech-to-speech models that reason through math without transcribing audio first, posting a large jump…