oxr
oxr is An open-source document parsing model with SOTA accuracy and speed, using an RSI loop during post-training. - SoMarkAI/OXR. It is ranked #1066 on the AI Radar, in RAG & memory, first seen 14 days ago and shared in 3 posts (25.4k views).
GitHub - SoMarkAI/OXR: An open-source document parsing model with SOTA accuracy and speed, using an RSI loop during post-training. · GitHub Skip to content Navigation Menu Sign in Search / Sign in Sign up Appearance settings You signed in with another tab or window. Reload to refresh your session. You signed out in another tab or window. Reload to refresh your session. You switched accounts on another tab or window. Reload to refresh your session. Dismiss alert {{ message }} Repository navigation OXR Public Notifications You must be signed in to change notification settings Fork 239 ( 239 ) Star 3.6k ( 3.6k ) You must be signed in to star a…
What people said about oxr on X
💥惊了！中国国产1.3B开源模型OXR横扫全球三大文档解析榜单！精度、速度双SOTA！ OmniDocBench 97.18、olmOCR-Bench 87.84、DocParsingBench 87.76三榜全第一，速度还快过Paddle和MinerU。 核心亮点： 🔹RSI递归自我改进：不依赖新增人工标注，自己找错、归因聚类、定向生成数据回流，哪儿薄弱补哪儿。 🔹架构重构：版面分析独立可插拔，主模型同时做识别+阅读顺序，彻底抹平性能损耗。 🔹13类元素全覆盖：表格、公式、化学式、代码块等，输出带坐标Markdown/JSON。 🔹部署友好：免注册免Key免费API + 本地部署，RTX 3060就能跑。…
— @NFT_Chen, 4 days ago · 21 likes · see the post
📑 真狠，SoMark 开源的 OXR 能把 PDF 和扫描件直接转成带版面坐标的 Markdown 或 JSON。 OXR 仓库 10 月 5 号才建，三周 GitHub 就涨到 1600 多 star，项目方称它在 olmOCR-Bench、OmniDocBench、DocParsingBench 三个基准上都排第一。 做过 RAG 或文档问答的都知道，最折腾的往往不是模型，而是前面那步解析：公式变乱码、表格拆成碎行、阅读顺序前后颠倒，后面检索怎么调都白搭。 OXR 是个 1.3B 参数的文档解析模型，从零预训练，公式、表格、化学结构式、代码块等 13…
— @Ryrenz, 9 days ago · 11 likes · see the post
Alternatives to oxr
- firecrawl.dev — @firecrawl - Give your AI agents web data and beyond
- Boo Bench — Leaderboard + paper
- auto-ontology — Auto Ontology 🔥
- liteparse — A fast, helpful, and open-source document parser. Contribute to run-llama/liteparse development by creating an account…
- arXiv.org — Contextual embeddings represent chunks of a document with respect to the document's global context. These models are…
- ModelBest — ModelBest — On-Device Open-Source Model Platform
oxr in numbers
- Rank on the AI Radar: #1066 of 10412
- Shared in 3 posts by 3 accounts: @NFT_Chen, @Ryrenz
- 25.4k views on those posts
- First seen 14 days ago, last shared 4 days ago
- Pricing seen by AGK Intelligence: open source
- Market: RAG & memory
FAQ
What is oxr?
An open-source document parsing model with SOTA accuracy and speed, using an RSI loop during post-training. - SoMarkAI/OXR It was first shared on X 14 days ago and is ranked #1066 on the AI Radar.
Is oxr free?
It is open source.
Who shared oxr?
3 accounts on X, including @NFT_Chen, @Ryrenz, in 3 posts totalling 25.4k views.
Every post is read and classified by AGK Intelligence: what it is, which market it belongs to, and whether the link is a real tool. 65.5k posts from 5k X accounts over the last 14 days, 10.4k tools, 19 markets. New posts are collected several times an hour and the full ranking is recomputed every hour — last update 2026-11-06 16:50 UTC. Full method.