🚨 AI News | TestingCatalog@testingcatalog

2026-06-24 00:33·9天前

AI 摘要

Mistral AI 发布 OCR 4 模型。在独立标注员对 600+ 现实文档（12+ 语言）的盲测中，OCR 4 被偏好，平均胜率 72%；OlmOCRBench 得分 85.20。OCR 4 还返回边界框、类型化块分类和行内置信度分数，作为 Search Toolkit 的组件，支持 170 种语言，且足够紧凑可单容器运行。

Mistral AI launched OCR 4 👀

Win rates averaging 72%， alongside the top overall score on OlmOCRBench （85.20）. Alongside the extracted text， OCR 4 returns bounding boxes， typed-block classification， and inline confidence scores. OCR 4 is an ingestion component of Search Toolkit， Mistral's open-source， composable search framework. Support for 170 languages across 10 language groups. OCR 4 is compact enough to run in a single container.

Mistral AIWe ran OCR 4 head-to-head against the field. Independent annotators blindly ranked 600+ real-world documents across 12+ languages, and preferred OCR 4 over ever...

多模态模型发布

在 X 查看原推导出 Markdown

🚨 AI News | TestingCatalog@testingcatalog · X

65导出 Markdown

2026-06-24 00:33·9天前

在 X 看原推· x.com

AI 摘要

Mistral AI launched OCR 4 👀

Win rates averaging 72%， alongside the top overall score on OlmOCRBench （85.20）. Alongside the extracted text， OCR 4 returns bounding boxes， typed-block classification， and inline confidence scores. OCR 4 is an ingestion component of Search Toolkit， Mistral's open-source， composable search framework. Support for 170 languages across 10 language groups. OCR 4 is compact enough to run in a single container.

Mistral AI