SoMark open-sources OXR, a 1.3B document parsing model with reported benchmark leads
OXR强势开源！1.3B横扫全球三榜第一，速度快过Paddle、MinerU，RSI让它卷出新高度
- Who
- SoMark team
- What
- released the OXR document parsing model, code, and weights
- When
- Quote
OXR 正式开源：主模型仅 1.3B 参数，在 OmniDocBench、olmOCR-Bench 和 DocParsingBench 三项基准测试中全部登顶，完整解析速度也超过了同场参测的 PaddleOCR-VL-1.6 和 MinerU2.5-Pro，把轻量模型的精度与效率同时推到了榜首。
What happened
SoMark has open-sourced OXR, a document parsing model with a 1.3B main model, according to a team-authored article republished by QbitAI. The team reports leading scores of 97.18, 87.84, and 87.76 on OmniDocBench, olmOCR-Bench, and DocParsingBench, respectively, and faster full parsing than PaddleOCR-VL-1.6 and MinerU2.5-Pro under the same test conditions. OXR handles 13 document element types, outputs Markdown / JSON with layout coordinates, and offers replaceable layout analysis; code, weights, an online demo, and a free API are available.
Read next
See the full storySoMark Open-Sources OXR Document Parsing Model, Reports Benchmark LeadsThe story so far
SoMark has open-sourced OXR, a document parsing system with a 1.3B-parameter main model, releasing its code and model weights alongside an online demo and a free API. OXR handles 13 types of document elements and produces Markdown or JSON with layout coordinates. Its layout analysis component can be replaced. In a team-authored article republished by QbitAI, SoMark reported benchmark-leading scores of 97.18 on OmniDocBench, 87.84 on olmOCR-Bench, and 87.76 on DocParsingBench. The team also said OXR completed full document parsing faster than PaddleOCR-VL-1.6 and MinerU2.5-Pro under the same test conditions.
Keep up with AI, one email a day
Model releases, price changes and compute developments, with original sources. Free, confirmed by email, and easy to unsubscribe. Read a past issue · 2026-10-27