This entry records what we found when we checked Block’s Dessa page on October 26, 2026. It is a status check, not a launch announcement.
What Block says Dessa is. A full-duplex, speech-to-speech conversation model built by a small AI research lab at Block, trained from scratch on recorded human conversation rather than text. Block says it listens and talks at the same time, uses short backchannel responses, and runs faster than real time.
What is available. Six featured recordings on Block’s page, each a live conversation between a person (or another AI) and Dessa. Block says more recordings will be shared on X.
What is not available yet. We found no public sign-up, app, download file, developer API, pricing, or release date. Block’s page says the model can run locally on a MacBook Pro M4 Max or better and shows a “Download” button, but that button is a placeholder tied to the note that launch links are coming soon. No download has been published.
Who can get early access. Block is offering early access to a small group of AI safety researchers. Its “get in touch” link opens a request form that asks for a name, an institutional email, and optionally a Google Scholar or blog link.