DeepSeek-OCR: Contexts Optical Compression
Haoran Wei et al.
arXiv:2510.18234
In short
DeepSeek-OCR tests whether text can be stored more compactly as an image: render a page, encode it into a small number of vision tokens, and decode the text back. At up to 10× compression it recovers about 97% of the text, and it is also a strong, efficient document reader.
Why it matters
It raises a new idea for long context: compress old history visually instead of keeping every text token.
Read first
The 3 Field Guide ideas this paper leans on.
Starting from scratch? The full route 13 ideas · basics first
- Machine Learning ✓ understood
Building systems that learn patterns from data instead of following hand-written rules, getting better at a task as they see more examples.
- Neural Network ✓ understood
A computational model inspired by biological neural networks, consisting of interconnected nodes (neurons) organized in layers that process information through weighted connections.
- Deep Learning ✓ understood
A subset of machine learning that uses neural networks with multiple layers (deep neural networks) to learn hierarchical representations of data.
- Computer Vision ✓ understood
The field of AI that gets computers to extract meaning from images and video: what is in them, where it is, and how it moves.
- Optical Character Recognition · read first ✓ understood
Converting images of text (scanned documents, photos) into machine-readable text using computer vision.
- Token · read first ✓ understood
The basic unit of text that a language model processes, typically representing a word, subword, or character. Tokens are the fundamental building blocks for LLM input and output.
- Dataset ✓ understood
A collection of data examples used for training, validating, or testing machine learning models.
- Feature ✓ understood
A single measurable property of an example, such as a house's floor area or how many links an email contains, used as an input to a model.
- Representation Learning ✓ understood
Learning useful features or representations of data automatically, rather than hand-crafting them.
- Embedding ✓ understood
A list of numbers (a vector) that represents a word, sentence, image or other item, learned so that similar items end up close together.
- Attention Mechanism ✓ understood
A technique that lets a neural network weigh every part of its input when producing each output, focusing on the parts most relevant at that step.
- Transformer ✓ understood
A neural network architecture, introduced in 2017, built from stacked self-attention and feed-forward layers; the basis of nearly every modern large language model.
- Context Window · read first ✓ understood
The maximum number of tokens an LLM can process at once, including both input prompt and generated output. Also called context length.
In the frontier
- Rank
- #28 of 100
- Citations
- 159
- as of Aug 9, 2026
- Published
- Oct 2025
Topics: Document AI and OCR , Efficiency and serving , Long context and attention
Selection: 1kpapers.com by Together AI, most-cited as of Aug 9, 2026