Google DeepMind 与 Artificial Analysis 联手推出 MKA，英文全称为 Multimodal Knowledge Agent。这个基准瞄准 Agent 处理视觉复杂任务的能力。
它把蓝图交叉比对、科学图表阅读等专业工作材料带进评测视野。Agent 面对的输入开始从文字问题，延伸到真实工作中的多模态文档。
原帖留影
With @ArtificialAnlys, we built MKA (Multimodal Knowledge Agent): a benchmark for how well agents handle visually complex tasks. Everyday professional work often uses more than just text – from cross-referencing blueprints to reading scientific charts. This is not covered by@GoogleDeepMind在 X 查看原帖发布于 2026/11/11 02:17（北京时间）