Unverified50% confidenceFactExact time
DSH Vision Toolkit 1.38（MIT）为纯文本模型提供图像问答、长截图 OCR、界面还原和像素差异等能力，并将图像证据持久保存以复用
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
8/27/2026
First Seen
Valid until: 11/25/2026
Sources
Related Entities
Related Claims
UnverifiedThe visual model receives each screenshot and outputs structured information, including OCR-recognized text, page layout, pixel coordinate bounding boxes of UI controls, and element states.71% similarUnverifiedDoubao 2.0 Mini supports Vision Understanding functionality, with the underlying layer using a Vision Transformer (ViT) architecture to encode images into token sequences.68% similarUnverifiedPawFlow 中视觉模型将屏幕截图转换为结构化语义描述，包括OCR文本、页面布局、UI控件的[x,y,w,h]像素坐标框及元素状态67% similarUnverified高分辨率视觉能力通过动态切片或高分辨率编码器实现，对代码截图分析、UI设计稿理解等场景重要66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/807693API
curl https://kongchang.com/api/v1/knowledge/claims/807693MCP
get_claim(id=807693)