Update: Gemini 2.5 Pro (I/O Edition) and Box AI
Blog post from Box
Google's early access release of Gemini 2.5 Pro Preview (I/O edition) marks a substantial upgrade from its predecessor, showing significant improvements in coding and interactive web app development capabilities. Box's evaluation of this version, particularly focusing on its reasoning capabilities for data extraction tasks, reveals notable performance gains. The I/O version of Gemini 2.5 Pro achieved an F1 score of 0.86 and 86.0% correctness on the CUAD dataset, compared to the initial version's F1 score of 0.82 and 80.0% correctness. Accuracy in multi-document Q&A also improved from 81.7% to 84.72%. Although the gains on the hard CUAD extraction subset and single-document Q&A were minimal, the improvements in handling extensive datasets and complex question-and-answer scenarios demonstrate its growing power in real-world business applications within the Box Intelligent Content Management platform. Gemini 2.5 Pro is available through Box AI Studio and Box AI APIs.
No tracked trend matches for this post yet.
Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.