DeepSeek V4 will be released next week and will have image and video generation capabilities, according to the Financial Times

Chinese AI company DeepSeek is preparing to launch its V4 model next week, marking a significant escalation in the global AI competition. According to Financial Times reporting, the new model will include image and video generation capabilities, transforming DeepSeek from a text-focused competitor into a full multimodal AI provider. This release represents China's most direct challenge yet to US AI dominance, coming at a time when geopolitical tensions around AI development are intensifying. The timing suggests DeepSeek aims to capture market share while Western companies face regulatory scrutiny and compute constraints.

Technically, DeepSeek V4's addition of image and video generation capabilities means it can now compete directly with OpenAI's DALL-E and Sora systems, as well as Anthropic's multimodal offerings. The model reportedly maintains DeepSeek's reputation for strong reasoning at lower costs, potentially offering enterprises an alternative to expensive US models. This development could accelerate AI adoption in Asia and emerging markets where cost sensitivity is higher, while also providing geopolitical diversification for global companies concerned about US export controls. The release will test whether Chinese AI models can match or exceed Western benchmarks while operating under different data and regulatory environments.