DeepSeek Unveils Multimodal LLM Aiming for Opus 4.8 Performance

2026-08-27

Chinese AI startup DeepSeek has launched V4 Flash Vision Exp, a new multimodal large language model. The model is currently accessible through DeepSeek's paid developer platform, with a free version potentially to follow.

Source: SiliconANGLE

Reported by VERA Newswire.