Jerry Liu 分享 GPT-6.1 Sol 文档 OCR 测试结果
我们对 GPT-6.1 Sol 的文档 OCR 任务表现做了全面测试。
相比一周前的 GPT-6 Sol（没错，才一周！），它在表格解析能力和阅读顺序上有明显提升，令人惊喜。GPT-6.1 Sol 的表格解析能力与 GPT-6 Astra 相近。
这些结果非常不错。前沿模型通常仍比高性价比的文档解析方案贵一个数量级，但这也意味着，在较高价位的一端（每页超过 1 美分）提升能力，同样还有很大的价值空间。
查看 ParseBench：https://www.parsebench.ai/
若想了解完整的文档解析能力，请查看 LlamaParse：https://cloud.llamaindex.ai/
对照原文
we comprehensively benchmarked gpt-6.1 sol on document OCR tasks. compared to gpt-6 sol from a week ago (!), we see a sizable increase in table parsing capabilities and reading order, which is pleasantly surprising. gpt-6.1 sol has similar table parsing as gpt-6 astra. these are great results. frontier models are still typically an order of magnitude more expensive than cost-effective document parsing solutions, but this also means that there's a lot of alpha in improving capabilities at the "premium" end (>1c per page) of the spectrum too check out ParseBench: https://t.co/PWczfhp0OX for a comprehensive set of document parsing capabilities, check out llamaparse: https://t.co/XYZmx5TFz8