Listeners, Chinese AI powerhouse DeepSeek is making waves with its highly anticipated V4 model, teased since January and now set for launch in the coming weeks. According to the US-China Economic and Security Review Commission's China Bulletin from April 2, 2026, V4 promises to outperform rivals like Anthropic's Claude and OpenAI's ChatGPT, especially in coding and long-context software engineering tasks, with internal tests beating Claude 3.5 Sonnet and GPT-4o on benchmarks.

The Information reports on April 3, 2026, that V4 will run entirely on Huawei chips, a major step for China's AI independence amid US export controls. DeepSeek collaborated with Huawei and Cambricon to port the model, excluding Nvidia from early access. Huawei's new Ascend 950PR chip, claimed to offer 2.8 times the power of Nvidia's H20, has seen massive orders from Alibaba, ByteDance, and Tencent—hundreds of thousands of units—driving prices up 20 percent, though production bottlenecks persist.

V4 is rumored to be multimodal, handling pictures, video, and text generation, upgrading from prior text-only versions, as noted in the China Bulletin. Huawei Central and Economic Times confirm the full reliance on Huawei AI chips, highlighting China's push to localize hardware.

However, controversy brews over training methods. The China Bulletin and a February Reuters memo reveal OpenAI accusing DeepSeek of "distillation attacks"—using fraudulent accounts to scrape outputs from US models like Claude and ChatGPT, bypassing safeguards despite China's domestic chip focus.

This blend of innovation and tension underscores the global AI race. Thank you for tuning in, listeners—come back next week for more. This has been a Quiet Please production, and for more, check out Quiet Please Dot A I.

Some great Deals https://amzn.to/49SJ3Qs

For more check out http://www.quietplease.ai

This content was created in partnership and with the help of Artificial Intelligence AI