Nvidia is expanding hardware and software support for Chinese open-weight AI models, including DeepSeek V4 Flash and Alibaba's Qwen 3.8, to keep developers within its technology ecosystem as Washington considers tighter restrictions on China-linked AI. In its latest SEC filing, the chipmaker warned that limits on corporate support for applications and services built on models such as DeepSeek, Qwen and Kimi could create a significant operational risk and materially affect its business. Nvidia's strategy includes day-zero support for Qwen 3.8 on RTX GPU systems and software updates that improve deployment of models such as Z.ai's GLM 5.2 and DeepSeek on DGX Spark systems. The company and more than 20 other firms, including Microsoft, Meta and Palantir, have urged policymakers to avoid premature restrictions on open-weight models. Nvidia shares rose about 7.6% Thursday after stronger-than-expected second-quarter results and revenue guidance above Wall Street estimates, and the stock has gained more than 10% so far this year.