Local AI Picksdata checked 2026-10-11

Local OCR and document parsing models

10 current open-weight models from 8 labs. Latest version of each family only.

Models

10
ModelLabParamsYour deviceScore
Youtu Parsing OmniTencent5.33B—LightOnOCR 3LightOn1.01B—WeVisDocTencent2.44B—Jina OCR v1Jina AI3.37B—ArmorOCRAnt Group inclusionAI8.77B—Nemotron Parse 2.0NVIDIA903M—Unlimited OCRBaidu3.34B—PP OCR v6Baidu PaddlePaddle——PaddleOCR VL 1.6Baidu PaddlePaddle959M—Granite Vision 4.1IBM4B—

Newest first.

Min memory is the smallest listed option. ● stated by the publisher · ◇ estimate from file size, an 8K-token context and 0.5 GB for the runtime. — means no figure is available.