# Block Unveils Dessa, a Full-Duplex Speech Model Trained Only on Speech

> **Publication**: aitalky (Independent AI Journalism)
> **Canonical Citation URL**: https://aitalky.vercel.app/news/block-unveils-dessa-a-full-duplex-speech-model-trained-only-on-speech
> **Original Source**: [MarkTechPost](https://www.marktechpost.com/2026/10/29/block-dessa-full-duplex-speech-model/)
> **Author**: Michal Sutter (undefined)
> **Published Date**: Thursday, October 29, 2026
> **Category**: RESEARCH
> **Tags**: AI, MarkTechPost, research

## Executive Summary
Block's AI lab released Dessa, a full-duplex speech-to-speech model trained only on human conversation audio. It runs locally on a MacBook Pro M4 Max, but Block has published no size, benchmarks or weights



## Full Coverage
Block's AI lab released Dessa, a full-duplex speech-to-speech model trained only on human conversation audio. It runs locally on a MacBook Pro M4 Max, but Block has published no size, benchmarks or weights

---
*Generated for AI agents and LLM citation by aitalky (https://aitalky.vercel.app).*