Science & Tech · AI · 8 hrs ago
Block unveils Dessa, a voice AI model trained on human speech
Block introduced Dessa, a new speech-to-speech AI model built by its research team.
The model was trained on human speech, and Block says it can run on a MacBook Pro with an M4 Max chip or better.
Dessa can listen and speak at the same time, a feature known as full-duplex conversation.
Block says this approach can preserve emotion and timing in a person's voice, which it argues text-based voice systems can miss.
The model's emotional responses can sound lifelike, though demonstrations also showed some awkward exchanges.
Block has not opened Dessa to the public and plans to give a small group of AI safety researchers early access to test it.
Block announced Dessa, a voice AI model trained entirely on human speech.
The model can listen and speak at the same time and is designed to run locally on consumer-grade computers.
Block says Dessa can convey emotion and nuance in speech, though shared demonstrations also showed some awkward exchanges.
The company says it built Dessa to explore speech-based AI, improve human-AI understanding and address safety concerns.
Block says a small group of AI safety researchers will get early access; there is no public access at present.
- Who
- Block, the company led by Jack Dorsey, announced the model.
- What
- It unveiled Dessa, a speech-to-speech AI model trained on human speech that can talk and listen at the same time.
- When
- Block announced it on October 23, 2026.
- Where
- Block’s research lab developed the model. It can run locally on a MacBook Pro M4 Max or better.
- Why
- Block cited curiosity, better understanding between humans and AI, and safety. It says voice AI often loses emotional information when speech is converted to text.
This story does not have two clearly opposing sides.
Block announced Dessa and shared demonstrations of conversations with the model.
TechCrunch reported that Block had released the model from its research lab.
Block said a small group of AI safety researchers would receive early access.
- Model
- Dessa
- Training
- Entirely on human speech
- Operation
- Full-duplex: it can speak and listen at the same time
- Local use
- MacBook Pro M4 Max or better
- Early access
- A small group of AI safety researchers
3 sources
What people are saying
Interacting with current voice AI can just plain suck. It turns your voice into text and throws away an incredible amount of information text alone can’t carry. The most powerful AIs in the world can’t hear how you feel. We think that’s ridiculous, leads to a worse experience, and is potentially dangerous,
Babies learn to talk without ever seeing a word written down. We wanted to see if a model could too. It can. It’s just very hard.
Speech is the most natural interface to any intelligence, super or otherwise, and today’s voice AI wastes it. It flattens what we say into text and can’t keep up with human timing.
AI keeps gaining power and independence, and it’s still nearly deaf to the emotion in our voices. Imagine being in serious pain while a voice AI has no idea and just keeps going. That’s the safety problem. Dessa is a step toward fixing it.
Posts on X
We made the first AI model you can have a realistic conversation with that was trained entirely on speech! Dessa is a full-duplex speech-to-speech model that you can run on consumer GPUs. This is a GPT3 moment for speech.
We believe voice is the most natural way to interact with intelligence, so we created Dessa, the most authentic full-duplex speech model ever built. Talking with it feels like talking with a person…full of emotion, nuance, pauses, and a natural cadence.
Meet Dessa, our AI model trained from scratch on nothing but speech. We believe speech is the natural interface for technology. So we built Dessa. And we’re inviting AI safety researchers to explore Dessa. Learn more: https://t.co/6wA8KhWdjn