China’s DeepSeek releases new AI model and claims it beats all open-source competitors
DeepSeek has released its new AI model, DeepSeek-V4, claiming it outperforms all open-source competitors. The model is available in two versions, DeepSeek-V4-Pro and DeepSeek-V4-Flash, with the latter being a more efficient and economical choice.
DeepSeek has released its new AI model, DeepSeek-V4, which it claims offers world-beating capabilities. The model is available in two versions: DeepSeek-V4-Pro and DeepSeek-V4-Flash. DeepSeek-V4-Pro significantly leads other open-source models in world knowledge benchmarks and is only slightly outperformed by Google's Gemini-Pro-3.1. The model features an ultra-long context of one million words and a “maximum reasoning effort mode”. DeepSeek-V4 is reportedly optimized for China's domestic chips and can process a maximum output of 384,000 tokens. The release comes amid growing US semiconductor export restrictions to China.
This content was automatically generated and/or translated by AI. It may contain inaccuracies. Please refer to the original sources for verification.