{"id":"oxr-1-0-somark","slug":"oxr-1-0-somark","name":"OXR-1.0","orgId":"somark","family":"oxr-1-0-somark","releaseDate":null,"availability":"announced","openWeights":true,"license":null,"paramsTotal":null,"paramsSource":null,"contextWindow":null,"knowledgeCutoff":null,"modelCardUrl":"https://modelscope.cn/models/SoMark/OXR-1.0","hfRepo":null,"modalities":["image","text"],"aliases":["OXR-1.0","oxr-1.0","oxr 1.0","somark oxr","somark oxr-1.0","somarkai/oxr-1.0","somark/oxr-1.0","optical everything recognition"],"aliasNote":null,"origin":"tracked","tracked":{"trackedSince":"2026-10-21","announced":"2026-10-21","blurb":"SoMark's open-weight document parsing model. It turns PDFs and document images into Markdown or JSON with layout coordinates, recognising 13 element types including tables, formulas, chemical equations and structures, stamps and code blocks. Layout analysis is a separate, swappable model. Weights under SoMark's OXR License, which adds commercial thresholds to Apache 2.0.","api":"live","openWeights":"yes","params":{"text":"1.3B (SoMark); the Hugging Face safetensors file totals about 1.53B","source":"SoMark's model card and Hugging Face metadata"},"facts":[{"label":"Task","value":"Document parsing (PDF and images to Markdown / JSON), with a pluggable layout-analysis model (D-FINE based, shipped with the weights)"},{"label":"Training","value":"Pretrained from scratch; post-training uses a recursive self-improvement loop on failed and weak samples"},{"label":"License","value":"OXR License: Apache 2.0 plus overriding terms. Commercial use is free only below USD 2M total funding and USD 2M annual revenue; online services must credit SoMark OXR; competing hosted parsing/OCR APIs are not allowed"},{"label":"Weights","value":"Hugging Face (SoMarkAI/OXR-1.0) and ModelScope (SoMark/OXR-1.0), BF16; repos created Oct 2–5, 2026"},{"label":"Hosted access","value":"Free API and online demo with no key or sign-up, no stability guarantee"},{"label":"Serving","value":"vLLM via an out-of-tree plugin; 24 GB GPU minimum, 48 GB+ recommended"}],"claims":["Lab claim: first in SoMark's own comparisons on olmOCR-bench (87.84), OmniDocBench (97.18) and DocParsingBench (87.76), with peer scores taken from published papers, model cards and leaderboards.","Lab claim: fastest of 21 models in SoMark's speed test at 557.77 native decode tokens/s and 0.99 pages/s on one RTX PRO 6000 Blackwell GPU."],"sources":[{"kind":"repo","title":"OXR-1.0 on ModelScope","url":"https://modelscope.cn/models/SoMark/OXR-1.0","date":"2026-10-21"},{"kind":"repo","title":"OXR-1.0 on Hugging Face","url":"https://huggingface.co/SoMarkAI/OXR-1.0","date":"2026-10-21"},{"kind":"repo","title":"OXR on GitHub (code, license and benchmark notes)","url":"https://github.com/SoMarkAI/OXR","date":"2026-10-21"},{"kind":"press","title":"ModelScope announcement on X","url":"https://x.com/ModelScope2022/status/2112901338178687182","date":"2026-10-21"}],"firstBenchmark":null},"organization":{"id":"somark","name":"SoMark","slug":"somark","country":null,"url":"https://somark.ai","color":"#11110F","mark":"S"},"price":null,"answer":{"word":"Pending benchmarks.","scope":"Announced Oct 21, 2026; no independent results yet.","lead":"As of Oct 22, 2026, OXR-1.0 is pending benchmarks on isitsota: announced Oct 21, 2026, with no independent results on the boards we track yet.","confidence":"low"},"status":"PENDING_BENCHMARKS","verdicts":[]}