DeepSeek Unveils New Multimodal Model Challenging Anthropic's Opus-4.8!

DeepSeek has launched its experimental multimodal model, DeepSeek-V4-Flash-Vision-Exp, claiming performance near Anthropic’s Opus-4.8. It surpasses Opus-4.8 in three out of eleven benchmarks. The model enhances text capabilities and integrates image understanding. Despite trailing in eight benchmarks, its competitive pricing may attract enterprise buyers. The release is generating excitement in the AI community.