DeepSeek-OCR
Read documents like an image. DeepSeek-OCR is a model that compresses long text by treating it as an image.
Reactions (1)
Similar products
by description similarity
Mistral OCR
mistral.aiIntroducing the world’s best document understanding API
Our open image-to-markdown OCR model — turning documents into structured text with tables, equations, and reading order preserved.
DeepSeek-VL2
github.comMoE vision-language, now easier to access
DocsRouter
docsrouter.comThe OpenRouter for OCR and Vision Models
Multi-modal OCR pipeline optimized for ML training (text, figure, math, tables, diagrams)