DE
DeepSeek-OCR 2
Release date: 2026-01-28Updated: 2026-06-14 23:13:18.699644
Parameters
350M
Context length
32K
Chinese support
Supported
Reasoning ability
DeepSeek-OCR 2 is an AI model published by DeepSeek-AI, released on 2026-01-28, for Multimodal model, with 350M parameters, and 32K context length, requiring about 7GB storage.
Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology
DeepSeek-OCR 2
Model basics
Reasoning traces
Not supported
Thinking modes
Thinking modes not supported
Context length
32K tokens
Max output length
8K tokens
Model type
Multimodal model
Modality (in / out)
Text, Image → Text
Release date
2026-01-28
Model file size
7GB
MoE architecture
Yes
Total params / Active params
350M / N/A
Knowledge cutoff
No data
DeepSeek-OCR 2
Open source & experience
Code license
Weights license
DeepSeek License Agreement- Commercial use permitted
Live demo
No live demo
DeepSeek-OCR 2
Official resources
Paper
DataLearnerAI blog
No blog post yet
DeepSeek-OCR 2
API details
API speed
4/5
No public API pricing yet.
DeepSeek-OCR 2
Benchmark Results
No benchmark data to show.
Compare with other models
No curated comparisons for this model yet.
Want a custom combination? Open the compare tool
DeepSeek-OCR 2
Publisher
DeepSeek-AI
View publisher details DeepSeek-OCR 2
Model Overview
DeepSeek-OCR 2 is an AI model published by DeepSeek-AI, released on 2026-01-28, for Multimodal model, with 350M parameters, and 32K context length, requiring about 7GB storage.
DataLearner on WeChat
Follow DataLearner on WeChat for AI model updates and research notes.
