---
title: "AlphaSignal Image: Ocr News"
url: "https://alphasignal.ai/topic/image/ocr"
description: "Latest AI and machine learning news about Ocr (Image) on AlphaSignal."
---

# AlphaSignal Image: Ocr News

Latest AI and machine learning news about Ocr (Image) on AlphaSignal.

## Articles

- [Umi-OCR Hits 47K Stars Bringing Private Offline Text Extraction to Desktops](https://alphasignal.ai/news/umi-ocr-hits-47k-stars-bringing-private-offline-text-extraction-to-desktops.md) (2026-10-10)
- [LightOn Releases LightOnOCR-3 to Parse Documents Without Extra Tools](https://alphasignal.ai/news/lighton-releases-lightonocr-3-to-parse-documents-without-extra-tools.md) (2026-10-07)
- [StarDoc-AI's TeleOCR Beats Gemini 3 Pro at Document Parsing With 1.2B Parameters](https://alphasignal.ai/news/stardoc-ai-s-teleocr-beats-gemini-3-pro-at-document-parsing-with-1-2b-parameters.md) (2026-09-26)
- [Sarvam AI's Vision 2.1 Beats Gemini at Reading India's 22 Languages](https://alphasignal.ai/news/sarvam-ai-s-vision-2-1-beats-gemini-at-reading-india-s-22-languages.md) (2026-09-24)
- [Jina AI's jina-ocr-v1 Parses PDF Pages at 2.57 Pages per Second](https://alphasignal.ai/news/jina-ai-s-jina-ocr-v1-parses-pdf-pages-at-2-57-pages-per-second.md) (2026-09-17)
- [Tencent's WeVisDoc-4B Tops Document Parsing Benchmarks Without OCR or Detectors](https://alphasignal.ai/news/tencent-s-wevisdoc-4b-tops-document-parsing-benchmarks-without-ocr-or-detectors.md) (2026-09-16)
- [Cohere Labs' Aya Vision 8B Beats Models 10x Its Size Across 23 Languages](https://alphasignal.ai/news/cohere-labs-aya-vision-8b-beats-models-10x-its-size-across-23-languages.md) (2026-07-09)
- [DeepSeek Releases DeepSeek-OCR to Process 200K Pages Daily on One GPU](https://alphasignal.ai/news/deepseek-releases-deepseek-ocr-to-process-200k-pages-daily-on-one-gpu.md) (2026-07-04)
- [Tencent's HunyuanOCR Beats Gemini-2.5-Pro at Document Parsing With 1B Parameters](https://alphasignal.ai/news/tencent-s-hunyuanocr-beats-gemini-2-5-pro-at-document-parsing-with-1b-parameters.md) (2026-07-01)
- [Baidu's Unlimited-OCR Parses 200-Page PDFs With Flat Memory Usage](https://alphasignal.ai/news/baidu-s-unlimited-ocr-parses-200-page-pdfs-with-flat-memory-usage.md) (2026-06-28)
- [Datalab's Chandra OCR 2 Beats GPT-4o on Documents at Half the Size](https://alphasignal.ai/news/datalab-s-chandra-ocr-2-beats-gpt-4o-on-documents-at-half-the-size.md) (2026-06-26)
- [Mistral OCR 4 Beats Every Rival With 72% Human Win Rate](https://alphasignal.ai/news/mistral-ocr-4-beats-every-rival-with-72-human-win-rate.md) (2026-06-23)
- [Baidu's Unlimited-OCR Reads Entire Documents Without Chunking or Stitching](https://alphasignal.ai/news/baidu-s-unlimited-ocr-reads-entire-documents-without-chunking-or-stitching.md) (2026-06-23)
- [Baidu's Unlimited-OCR Shrinks to 4-Bit to Fit on One GPU](https://alphasignal.ai/news/baidu-s-unlimited-ocr-shrinks-to-4-bit-to-fit-on-one-gpu.md) (2026-06-23)
- [PaddleOCR's PP-OCRv6 Beats GPT-5.5 and Qwen3-VL-235B at 34.5M Parameters](https://alphasignal.ai/news/paddleocr-s-pp-ocrv6-beats-gpt-5-5-and-qwen3-vl-235b-at-34-5m-parameters.md) (2026-06-10)
- [Comic Translate v2.8.0 Translates Manga 5x Faster With Half the Size](https://alphasignal.ai/news/comic-translate-v2-8-0-translates-manga-5x-faster-with-half-the-size.md) (2026-06-06)
- [iOS-OCR-Server Turns Your Idle iPhone Into a Free Private OCR Engine](https://alphasignal.ai/news/ios-ocr-server-turns-your-idle-iphone-into-a-free-private-ocr-engine.md) (2026-06-04)
- [Datalab's Surya 2 Beats 9B Models at OCR With Just 650M Parameters](https://alphasignal.ai/news/datalab-s-surya-2-beats-9b-models-at-ocr-with-just-650m-parameters.md) (2026-06-02)
- [Baidu's PaddleOCR-VL-1.6 Beats 235B Models at Document Parsing With Just 0.9B](https://alphasignal.ai/news/baidu-s-paddleocr-vl-1-6-beats-235b-models-at-document-parsing-with-just-0-9b.md) (2026-05-29)
- [Baidu's PaddleOCR-VL-1.6 Beats 235B Giants at Document Parsing With 1B Parameters](https://alphasignal.ai/news/baidu-s-paddleocr-vl-1-6-beats-235b-giants-at-document-parsing-with-1b.md) (2026-05-27)
- [Datalab's Surya OCR 2 Beats 9B Models at a Fraction of the Size](https://alphasignal.ai/news/datalab-s-surya-ocr-2-beats-9b-models-at-a-fraction-of-the-size.md) (2026-05-14)
- [Japan's National Diet Library Ships NDLOCR-Lite to Read Handwritten Japanese Without a GPU](https://alphasignal.ai/news/japan-s-national-diet-library-ships-ndlocr-lite-to-read-handwritten-japanese.md) (2026-04-24)
- [Fudan's OpenOCR Beats 7B Document Parsers With a Tiny 0.1B Model](https://alphasignal.ai/news/fudan-s-openocr-beats-7b-document-parsers-with-a-tiny-0-1b-model.md) (2026-04-23)
- [Zhipu AI's GLM-OCR Beats Gemini and GPT-5.2 at 260x Smaller](https://alphasignal.ai/news/zhipu-ai-s-glm-ocr-beats-gemini-and-gpt-5-2-at-260x-smaller.md) (2026-04-21)
- [GlotOCR Bench Reveals GPT-4.1 Fails on 92% of the World's Scripts](https://alphasignal.ai/news/glotocr-bench-reveals-gpt-4-1-fails-on-92-of-the-world-s-scripts.md) (2026-04-14)
- [Frankfurt Builds VisualStemma to Trace Manuscripts Without Transcription](https://alphasignal.ai/news/frankfurt-builds-visualstemma-to-trace-manuscripts-without-transcription.md) (2026-04-13)
- [RapidOCR Runs PaddleOCR Models Anywhere Without the Heavy Framework](https://alphasignal.ai/news/rapidocr-runs-paddleocr-models-anywhere-without-the-heavy-framework.md) (2026-04-11)
- [Alibaba's Logics-Parsing-v2 Beats Every Open-Source OCR Model on Complex Documents](https://alphasignal.ai/news/alibaba-s-logics-parsing-v2-beats-every-open-source-ocr-model-on-complex.md) (2026-04-08)
- [Xiaohongshu's dots.ocr Beats Gemini 2.5 Pro on Multilingual Parsing at 3B Parameters](https://alphasignal.ai/news/xiaohongshu-s-dots-ocr-beats-gemini-2-5-pro-on-multilingual-parsing-at-3b.md) (2026-03-24)
- [Datalab's Chandra OCR 2 Beats Gemini and GPT-5 Mini Across 43 Languages](https://alphasignal.ai/news/datalab-s-chandra-ocr-2-beats-gemini-and-gpt-5-mini-across-43-languages.md) (2026-03-23)
- [Baidu's Qianfan-OCR Beats Gemini and Qwen3-VL-235B With Just 4B Parameters](https://alphasignal.ai/news/baidu-s-qianfan-ocr-beats-gemini-and-qwen3-vl-235b-with-just-4b-parameters.md) (2026-03-18)
- [Z.ai's Tiny GLM-OCR Beats Models 260x Bigger at Document Parsing](https://alphasignal.ai/news/z-ai-s-tiny-glm-ocr-beats-models-260x-bigger-at-document-parsing.md) (2026-03-16)
- [DualTSR Sharpens Blurry Text 27% Better Without an External OCR Model](https://alphasignal.ai/news/dualtsr-sharpens-blurry-text-27-better-without-an-external-ocr-model.md) (2026-03-15)
- [Xiaohongshu's dots.mocr Turns Charts and Diagrams Into Editable SVG Code](https://alphasignal.ai/news/xiaohongshu-s-dots-mocr-turns-charts-and-diagrams-into-editable-svg-code.md) (2026-03-13)
- [Xiaohongshu's FireRed-OCR Beats Gemini and Qwen3-VL-235B at Document Parsing](https://alphasignal.ai/news/xiaohongshu-s-firered-ocr-beats-gemini-and-qwen3-vl-235b-at-document-parsing.md) (2026-03-12)
- [Bilibili's TextFlux Renders Multilingual Text in Images With 1% of Training Data](https://alphasignal.ai/news/bilibili-s-textflux-renders-multilingual-text-in-images-with-1-of-training-data.md) (2026-03-12)
- [Zhipu AI's Tiny GLM-OCR Beats Gemini 3 Pro at Document Parsing](https://alphasignal.ai/news/zhipu-ai-s-tiny-glm-ocr-beats-gemini-3-pro-at-document-parsing.md) (2026-03-11)
- [Baidu's Qianfan-OCR Beats 235B Models at Document Parsing With Just 4B](https://alphasignal.ai/news/baidu-s-qianfan-ocr-beats-235b-models-at-document-parsing-with-just-4b.md) (2026-03-11)
- [Z.ai's Tiny GLM-OCR Beats Models 260x Larger at Reading Documents](https://alphasignal.ai/news/z-ai-s-tiny-glm-ocr-beats-models-260x-larger-at-reading-documents.md) (2026-01-30)
- [Baidu's PP-DocLayoutV3 Reads Curved and Skewed Docs With 33M Parameters](https://alphasignal.ai/news/baidu-s-pp-doclayoutv3-reads-curved-and-skewed-docs-with-33m-parameters.md) (2026-01-20)
- [LightOn's LightOnOCR-2 Beats 9B Models at OCR While Running 3.3x Faster](https://alphasignal.ai/news/lighton-s-lightonocr-2-beats-9b-models-at-ocr-while-running-3-3x-faster.md) (2026-01-16)
- [Baidu's PaddleOCR-VL Beats Gemini 2.5 Pro at Document Parsing With 0.9B Parameters](https://alphasignal.ai/news/baidu-s-paddleocr-vl-beats-gemini-2-5-pro-at-document-parsing-with-0-9b.md) (2025-10-16)
- [OpenBMB's MiniCPM-V 4.5 Beats GPT-4o on Benchmarks Running on an iPad](https://alphasignal.ai/news/openbmb-s-minicpm-v-4-5-beats-gpt-4o-on-benchmarks-running-on-an-ipad.md) (2025-08-24)
- [Xiaohongshu's dots.ocr Beats Gemini 2.5 Pro on Documents With a Tiny 1.7B Model](https://alphasignal.ai/news/xiaohongshu-s-dots-ocr-beats-gemini-2-5-pro-on-documents-with-a-tiny-1-7b-model.md) (2025-07-30)
- [Nanonets-OCR-s Turns Messy Documents Into Structured Markdown on a Single GPU](https://alphasignal.ai/news/nanonets-ocr-s-turns-messy-documents-into-structured-markdown-on-a-single-gpu.md) (2025-06-10)
- [IBM Research's Heron Boosts Open-Source PDF Parsing Accuracy by 23.5%](https://alphasignal.ai/news/ibm-research-s-heron-boosts-open-source-pdf-parsing-accuracy-by-23-5.md) (2025-04-15)
- [IBM's SmolDocling Beats 7B Models at Document Parsing With Just 256M Parameters](https://alphasignal.ai/news/ibm-s-smoldocling-beats-7b-models-at-document-parsing-with-just-256m-parameters.md) (2025-02-12)
- [NVIDIA's NVLM-D-72B Beats GPT-4o on OCR While Boosting Its Own Text Skills](https://alphasignal.ai/news/nvidia-s-nvlm-d-72b-beats-gpt-4o-on-ocr-while-boosting-its-own-text-skills.md) (2024-09-30)
- [StepFun's GOT-OCR2.0 Replaces Entire OCR Pipelines With One 580M Model](https://alphasignal.ai/news/stepfun-s-got-ocr2-0-replaces-entire-ocr-pipelines-with-one-580m-model.md) (2024-09-12)
- [OpenBMB's MiniCPM-Llama3-V 2.5 Beats GPT-4V on OCR Running on a Smartphone](https://alphasignal.ai/news/openbmb-s-minicpm-llama3-v-2-5-beats-gpt-4v-on-ocr-running-on-a-smartphone.md) (2024-05-19)