{"version":1,"type":"story","url":"https://digestai.news/story/block-releases-dessa-a-fullduplex-speech-model-trained-only-on-speech","json":"https://digestai.news/story/block-releases-dessa-a-fullduplex-speech-model-trained-only-on-speech.json","markdown":"https://digestai.news/story/block-releases-dessa-a-fullduplex-speech-model-trained-only-on-speech.md","slug":"block-releases-dessa-a-fullduplex-speech-model-trained-only-on-speech","headline":"Block releases Dessa, a full‑duplex speech model trained only on speech","summary":"Block’s AI research lab announced Dessa on October 23, 2026 as a research preview. The model is a full‑duplex speech‑to‑speech system that listens and talks simultaneously, trained solely on human conversation audio with no text or pretrained models. Block says it learned from real conversation cues such as backchannels, mid‑sentence laughs, and natural interruptions.\n\nDessa was trained on a single NVIDIA GB200 NVL72 rack, which links 72 Blackwell GPUs and 36 Grace CPUs. The lab claims the model runs faster than real time on a MacBook Pro M4 Max or better. No benchmark scores, weights, paper, API or price are released, and access is currently via a waitlist.\n\nThe article notes that other full‑duplex models, such as Moshi and PersonaPlex, use different training pipelines and larger compute budgets. Block argues that a speech‑only approach preserves emotional and prosodic signals that are lost in text‑first pipelines.","keyPoints":["Block’s Dessa is a full‑duplex speech‑to‑speech model trained only on human conversation audio, with no text or pretrained models.","It was trained on a single NVIDIA GB200 NVL72 rack (72 GPUs) and runs faster than real time on a MacBook Pro M4 Max.","No benchmark scores, weights, or API are released; access is via a waitlist and the model is a research preview."],"whyItMatters":"Dessa shows that speech‑only training can produce conversational models that retain emotional cues, challenging the dominant text‑first pipeline. If it scales, it could reduce data and compute needs for voice agents and open new possibilities for privacy‑sensitive, offline speech applications.","category":{"slug":"models","name":"Generative AI & Models","url":"https://digestai.news/category/models"},"entities":{"companies":["Block","OpenAI"],"models":["Dessa","Moshi","PersonaPlex","GPT-Live-1"],"people":["Michal Sutter"]},"firstPublishedAt":"2026-10-29T07:50:39Z","updatedAt":"2026-10-29T07:50:39Z","sourceCount":1,"hasPrimarySource":false,"sources":[{"outlet":"MarkTechPost","title":"Block Unveils Dessa, a Full-Duplex Speech Model Trained Only on Speech","url":"https://marktechpost.com/2026/10/29/block-dessa-full-duplex-speech-model","publishedAt":"2026-10-29T07:50:39Z","type":"press","primary":false,"lead":true}],"sourceNotes":null,"discussions":[],"thread":{"title":"Block Unveils Speech-Only AI Model","url":"https://digestai.news/thread/block-releases-dessa-an-ai-model-trained-only-on-human-speech","storyCount":2},"cite":{"text":"Digest AI, \"Block releases Dessa, a full‑duplex speech model trained only on speech\", 29 October 2026, https://digestai.news/story/block-releases-dessa-a-fullduplex-speech-model-trained-only-on-speech","publisher":"Digest AI","title":"Block releases Dessa, a full‑duplex speech model trained only on speech","datePublished":"2026-10-29T07:50:39Z","url":"https://digestai.news/story/block-releases-dessa-a-fullduplex-speech-model-trained-only-on-speech"},"generatedBy":"Written by Digest AI's editorial model from the linked sources; the sources are the record.","license":"Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse"}