Liquid AI releases on-device decision models d1-3B and d1-omni-600M
The story so far
Liquid AI has released open-weight decision models d1-3B and experimental d1-omni-600M on Hugging Face. Unlike word-by-word generation, these two models directly select the answer in a single forward computation. d1-3B accepts text and image input, and the smaller experimental model also supports text and audio input. Liquid AI said that d1-3B takes 8 milliseconds on RTX 4090 and 50 milliseconds on Jetson Orin Nano.
AI-written from the reports · Updated 28 days ago
Coverage timeline
Follow the coverage to see every side of the story.
- Liquid AIFeaturedLiquid AI releases open-weight decision models for edge devices
Liquid AI released d1-3B and the experimental d1-omni-600M on Hugging Face. These models choose an answer in a single forward pass rather than generating a token sequence. d1-3B accepts text and images; the smaller experimental model also supports text-and-audio inputs. The company reports 8 ms for d1-3B on an RTX 4090 and 50 ms on a Jetson Orin Nano.