Unlimited-OCR
Vision modelOCR modelUnlimited-OCR
Parameters
3.336B
Context length
32K
Chinese support
Supported
Reasoning ability
Unlimited-OCR 是百度于 2026 年 6 月 22 日开源的 OCR / 文档解析视觉语言模型，采用 R-SWA 注意力机制支持 32K 长输出和多页文档一次性解析，模型约 3B 总参数、0.5B 激活参数，MIT 许可。
Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology
Unlimited-OCR
Model basics
Reasoning traces
Not supported
Thinking modes
Thinking modes not supported
Context length
32K tokens
Max output length
32K tokens
Model type
Vision model
Modality (in / out)
Text, Image → Text
Release date
2026-06-22
Model file size
6.67 GB BF16 safetensors
MoE architecture
Yes
Total params / Active params
3.336B / 500M
Knowledge cutoff
No data
Unlimited-OCR
Open source & experience
Unlimited-OCR
Official resources
DataLearnerAI blog
No blog post yet
Unlimited-OCR
API details
API speed
4/5
No public API pricing yet.
Unlimited-OCR
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
Unlimited-OCR
Publisher
Unlimited-OCR
Model Overview
Unlimited-OCR 是百度于 2026 年 6 月 22 日开源的 OCR / 文档解析视觉语言模型，采用 R-SWA 注意力机制支持 32K 长输出和多页文档一次性解析，模型约 3B 总参数、0.5B 激活参数，MIT 许可。
DataLearner on WeChat
Follow DataLearner on WeChat for AI model updates and research notes.
