Block launches full-duplex speech model Dessa, supporting local operation on MacBook
Block launches the full-duplex speech model Dessa, which can listen to the user's speech while speaking; emotional expression is natively built into the model, and it can simulate states such as laughter, pauses, and choking up. The company says the model can run locally on consumer-grade computers, with the device threshold being MacBooks with M4 Max or higher configurations, and that training used only a single rack of NVIDIA GPUs. At this stage Dessa is not open for public trial; Block is sending testing invitations to safety researchers, focusing on risk scenarios in which such speech agents deceive humans.
Why read this
AI score 62
Block's self-developed speech model opts for on-device operation and specifies the M4 Max as a concrete device threshold, offering insight into the cost and hardware requirements of localizing speech models.
Topics
Body · AI translation
Block Unveils Full-Duplex AI Voice Model Dessa: Highly Human-Like, Supports Local Running on Apple MacBook
English text is being prepared.
English text is being prepared.
Dessa is a full-duplex model, and like GPT-Live released by OpenAI earlier this year, it can listen to the user's voice while speaking. Mainstream voice models generally need to be paired with a large language model to retrieve knowledge; Block says Dessa's own knowledge comes from the conversational data used in its training.
Block says the existing voice AI experience is often stiff and jarring, and the root cause is that the generation pipeline first converts to text and then uses the text to drive speech synthesis. Many models, such as ElevenLabs v4 and Fish Audio Drama 3, rely on voice tags to control emotion. Dorsey's company claims that emotional expression is natively built into the model, though it is currently unclear whether it still relies on some kind of tags underneath to regulate emotion.
Block wrote in its blog: "Today's voice AI interaction experience is really bad. It first converts your voice into text, and a large amount of information that text cannot carry is simply lost. Even the world's strongest AI cannot hear your emotions. We think this is unreasonable—not only is the experience poor, but there are also potential risks."
The research team released several sample conversations with Dessa, and this agent can switch timbre, speaking rate, and emotion. For example, in one conversation, the AI says it is tired from work, and its vocal state simultaneously carries a sense of fatigue.
In another example, the AI talks about the sense of reassurance the other person brings it, and its tone is emotional, almost choking up. Another demo shows Block employees playing a game with this voice assistant.
According to the official explanation, the model can run locally on consumer-grade computers, with the device threshold being an M4 Max or higher-configuration MacBook. But at this stage Dessa is not yet open for public trial; Block is sending testing invitations to safety researchers, focusing on testing risk scenarios in which such voice agents deceive humans.
English text is being prepared.
English text is being prepared.