SenseTime Open-Sources SenseNova U1.5, an 8B Native Unified Model

SenseTime released the technical report and open-sourced SenseNova-U1.5, an 8B native unified model for image understanding, generation and editing that requires no external vision encoder or VAE, with full training details disclosed.

2026-09-16 ~ 2026-09-17 · 2 related posts

• SenseNova-U1.5 Technical Report: VAE-Free Native 4K Image Generation — Secret_Yak2496 · 2026-09-16
• SenseTime open-sources SenseNova U1.5: 8B unified model beating GPT-Image-2 on visual reasoning — liuziwei7 · 2026-09-17