LightOnOCR-3 发布：高性能轻量级 OCR 与布局提取模型
原标题：LightOnOCR-3: High-Performance OCR and Layout Extraction in One Model
AI 摘要
LightOnOCR-3 是 LightOn 发布的新一代轻量级 OCR 模型系列，包含 0.8B、1B 和 4B 三个尺寸，采用 Apache 2.0 许可。相比前代，新模型在速度和转录质量上有显著提升，并新增了带标签的边界框输出、图像描述以及图表数值提取等视觉理解能力。在 olmOCR-Bench 上，4B 模型总体得分 86.3，仅落后于参数量 35.1B 的 Infinity Parser Pro 1.3 分；在 ParseBench 上，4B 和 0.8B 变体分别位列前两名。
正文节选
We are releasing LightOnOCR-3 a new family of highly performant lightweight OCR models. Compared to the previous generation, they introduce significant improvements in speed and transcription quality while introducing new visual understanding features. LightOnOCR-3 models are now able to output bounding box coordinates with labels for all document regions, image descriptions as well as extract numerical data in figures or charts. With these new capabilities, the models offer a ready-to-use, easi