Chinese AI company DeepSeek is set to release its next-generation model, DeepSeek-V3-0526, imminently, with multiple reports indicating performance comparable to leading AI models such as GPT-4.5 and Claude 4 Opus. The release has generated significant demand, reportedly causing service outages. Following this, DeepSeek plans to launch DeepSeek R2 by the end of June, which is expected to achieve a 97% cost efficiency. Industry speculation suggests that DeepSeek R2 may not yet incorporate a hybrid approach to AI processing like some competitors but could represent a substantial upgrade. The model is believed to be trained on Huawei's Ascend GPUs, and its release could impact major AI hardware suppliers such as NVIDIA. The company’s aggressive pricing and technological advancements are viewed as a potential challenge to American AI labs, which might need to adopt higher profit margins to remain competitive.