Knowledge Work

Exploring the New Multimodal Knowledge Agent Benchmark for AI Models

The Multimodal Knowledge Agent benchmark evaluates AI models' reasoning across visually complex tasks, developed in partnership with domain experts.