Unverified50% confidenceFactExact time
Donut是无OCR依赖的端到端文档理解模型，Pix2Struct将网页截图作为预训练数据，均为专用文档视觉编码器
1
Sources
50%
Confidence
Long-term
Relevance
8/20/2026
First Seen
Sources
Related Claims
UnverifiedDonut（Document Understanding Transformer）无需OCR预处理即可端到端地从文档图像中提取结构化信息71% similarUnverifiedwebpack 构建的依赖图是一个有向无环图（DAG）62% similarUnverifiedOpenCLI通过读取结构化的DOM快照获取页面内容，而非依赖截图61% similarUnverifiedRendemo通过Chrome扩展录制产品操作，捕获页面真实的DOM/HTML结构而非截图60% similarUnverified通过新增的Site功能，用户可以将产出结果直接部署为一个可访问的网页59% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/780095API
curl https://kongchang.com/api/v1/knowledge/claims/780095MCP
get_claim(id=780095)