Latest / Frontier models and safety
Block unveils Dessa, a conversation model trained only on human speech
See how this story connectsA small research lab at Block that reports to Jack Dorsey announced Dessa, a full-duplex speech-to-speech model trained from scratch on human conversation, with no text and no pretrained models. Block says it listens and talks at the same time, hears and speaks emotion, runs faster than real time locally on a MacBook Pro M4 Max or better, and was trained on one NVIDIA GB200 NVL72. The public can join a waitlist, and a small group of AI safety researchers is getting early access to look for risks.
Why it matters
It is a test of whether a model can learn language and facts from sound alone, and Block is asking outside safety researchers to probe it before a wider release.
- When
- 23 Oct 2026
- ID
WTA-N-2026-0413- Where
- United States
- Who
- Block, NVIDIA; Jack Dorsey, Ragavan Thurairatnam
- Topics
- Model releases, Safety testing
Summary by wheresthe.ai, from Block and the reporting credited above. Checked against the source on 25 Oct 2026.
Source copy: download the page as saved on 25 Oct 2026 (PDF). Also held by the Internet Archive.
Cite this story
WTA-N-2026-0413: “Block unveils Dessa, a conversation model trained only on human speech.” wheresthe.ai, 23 Oct 2026. https://wheresthe.ai/d/us-block-unveils-dessa-speech-only-conversation-model-2026-10/ (source: Block).
Spotted an error? Email [email protected].