Physix Frontier badgePhysix Frontier
PublishChinese | EN

Physix Frontier · News Briefing Card (QbitAI · Sep 22, 2026)

SenseTime launches SenseNova U1 Pro with multi-round editing

KEY FACTS

  • SenseTime has officially launched the SenseNova U1 Pro image generation model, which was previously showcased at WAIC.
  • The model is now available on SenseTime's Xiaohuanxiong platform, with its API simultaneously opened to enterprise customers.
  • The model supports image generation starting at 2K and up to 4K resolution.
  • Hands-on testing shows the model can perform multiple consecutive rounds of local image edits without distortion.
  • The model can autonomously retrieve information based on prompts to generate educational infographic posters.

KEY DATA

4KMax resolution
2KMin resolution
3 roundsTested edit rounds

PHYSIX OBSERVATION

Competition in image generation models is shifting from single-image output quality to deliverability, with continuous refinement and autonomous retrieval being key thresholds for commercial deployment. SenseTime bundles multi-round editing and web-connected capabilities into U1 Pro, targeting real workflows in e-commerce and advertising, which is more compelling to enterprise customers than simply topping benchmarks.

Source: QbitAI report