In anticipation of the release, major Chinese tech companies—including Alibaba Group, ByteDance, and Tencent Holdings—have reportedly placed bulk orders for Huawei’s upcoming chips, totaling hundreds of thousands of units, citing sources familiar with the purchases.
The next-generation AI model is expected to launch within the next few weeks, though neither Huawei nor DeepSeek responded immediately to Reuters’ requests for comment outside regular business hours.
According to the report, DeepSeek has collaborated closely with Huawei and Cambricon Technologies over the past months to rewrite portions of the model’s underlying code. In testing, the model has been optimized specifically for Chinese chip architecture. DeepSeek is also reportedly developing two additional V4 variants, each tailored for different capabilities and designed to run exclusively on domestic chips.
Earlier this year, Reuters reported that DeepSeek had deviated from standard industry practice by not showing its flagship model to U.S. chipmakers for performance optimization, granting early access instead to domestic suppliers, including Huawei.
The release of DeepSeek’s low-cost models, V3 and R1, last year had triggered a global tech stock selloff, raising questions among investors about the necessity for U.S. AI firms to spend billions on computing infrastructure. Since then, anticipation has mounted around DeepSeek V4, a next-generation model that could further reshape the AI landscape.