Chinese AI lab DeepSeek has unveiled two preview versions of its newest large language model, DeepSeek V4, which it claims is a significant leap forward from last year’s V3.2 and R1 reasoning models.
The V4 Flash and V4 Pro are both mixture-of-experts models with vast context windows—capable of handling extensive codebases or documents in prompts. The Pro model boasts an impressive 1.6 trillion parameters, making it the largest open-source model available.
Despite its size, DeepSeek claims both models outperform their predecessors and some leading models on reasoning benchmarks. However, they lag slightly behind OpenAI’s GPT-5.4 and Google’s Gemini 3.1 Pro in knowledge tests, suggesting a “developmental trajectory that trails state-of-the-art frontier models by approximately 3 to 6 months.”
One of the standout features is DeepSeek V4's affordability. The smaller V4 Flash model costs $0.14 per million input tokens and $0.28 per million output tokens, undercutting several competitors including GPT-5.4 Nano and Claude Haiku 4.5.
However, the launch comes amid accusations from US authorities of China’s AI labs stealing IP on an industrial scale through proxy accounts, with DeepSeek itself facing similar allegations from Anthropic and OpenAI regarding model ‘distillation’.
