178 likesNew · 9d old
Full-duplex voice model lets AI converse as naturally as a call
What is Sophon-Realtime-v1?
Sophon-Realtime v1 is an open-source 32-billion-parameter model that can listen and speak at the same time. It picks up tone, hesitation and laughter, then replies with matching vocal cues, offering three builds for expressive speech, speed, or low-latency output.
Who it is for
Developers building voice assistants, chatbots or interactive characters can use it to create natural, back-and-forth spoken conversations.
How to try it
Open the demo on Hugging Face to test the model live.
The story so far
Early Signal first spotted adventists-ai/Sophon-Realtime-v1 on November 12, 2026, with 178 likes. On its busiest day it gained 29 likes. The project is 9 days old. Early Signal files it under Voice.
Questions about Sophon-Realtime-v1
- How do I try Sophon-Realtime-v1?
- Open the demo on Hugging Face to test the model live.
- What do I need to run Sophon-Realtime-v1?
- A single GPU with about 80 GB VRAM