✦
Summary
Dessa is Block's full-duplex conversation model, trained directly on human speech to hold expressive real-time conversations. It learns from audio without a text-first transcription pipeline or pretrained language backbone, preserving timing and vocal information that text alone cannot carry.
Key Features
Full-duplex spoken conversation.
Speech-only training.
Real-time listening and speaking.
Natural turn-taking.
Expressive vocal delivery.
Context-sensitive laughter.
Volume adjustment from instructions.
Research-preview waitlist.
About Dessa
It offers a research preview of voice interaction with natural interruptions, emotion, and cadence, with early access managed by waitlist. Core capabilities include Full-duplex spoken conversation, Speech-only training, Real-time listening and speaking, and Natural turn-taking.
Join the Dessa waitlist. The project provides recorded demonstrations but has not published a general commercial price or downloadable model. Additional characteristics include Expressive vocal delivery, Context-sensitive laughter, Volume adjustment from instructions, and Research-preview waitlist.