Unlimited OCR 2

Released
2026-10-27
Last refreshed
2026-10-27
Status
Researched today
Open sourceCommercial use: permittedMultimodalVisionVision

Unlimited OCR 2 is available now for vision with open-source; evaluate it while provider pricing coverage matures.

Baidu AI releases · 9 in the last 12 months · this family litChangelog →
Specifications
Released
2026-10-27
Parameters
3.3B
Architecture
Mixture of Experts
Specialization
ocr
Openness
Open source
License
MITOSI-approvedPermitted
Weights
Available
Code
Unknown
Created by
Baidu AI

Innovative text-to-video and app builder

Beijing, China
Founded 2010
Website
Pricing

No tracked provider pricing is available yet.

Links
Website HuggingFace

About

Unlimited OCR 2 is Baidu's open document-parsing model. It uses SR-SWA sliding-window attention so multi-page documents of virtually unlimited length can be parsed in one request without attention cost growing with length. A 3.3B-parameter mixture-of-experts vision-language model (64 routed experts, 6 active), it is released on Hugging Face and ModelScope under the MIT license.

Providers

No active provider routes are tracked for this model yet.

Capabilities

VisionMultimodal

Benchmark peer barsfor Vision

No task-mapped benchmark peers are available for this model yet.

Migration checks

No linked migration route is available for this model yet.

API versions

baidu/Unlimited-OCR-2PaddlePaddle/Unlimited-OCR-2Unlimited-OCR-2