SenseNova-U1-8B-MoT

Autoregressive understanding + flow matching generation model. Open source Nano Banana / GPT-Image-2-style model. Unified text-to-image and image editing with SenseNova-U1-8B-MoT, optionally accelerated by the official 8-step LoRA.

Mode
Aspect ratio (text-to-image only — editing keeps input ratio)
0 2147483647
Examples