U.S. AI startups race to develop affordable alternatives to Chinese models amid funding challenges

Huo Xing Cai Jing reports that on August 2, according to The Wall Street Journal, as Chinese open-weight models such as Kimi, Qwen, and DeepSeek approach the performance of top U.S. models at significantly lower costs, Silicon Valley and Washington are increasingly concerned that Chinese models may permanently compress the profitability of American AI companies. U.S. startups including Arcee AI, Reflection AI, and Poolside are racing to develop domestic alternatives to meet user demand for low-cost, downloadable, and customizable models. However, American open-weight model companies face difficulties in securing funding. Some investors question whether freely open models can generate stable revenue and worry that such technologies may undermine the value of their investments in OpenAI and Anthropic. In the first quarter of 2026, AI startups raised a total of $255.5 billion, with nearly two-thirds coming from just three funding rounds by OpenAI, Anthropic, and xAI. Despite limited funding, Arcee AI completed 33 days of pre-training using 2,048 NVIDIA Blackwell B300 chips on a budget of approximately $20 million, launching the Trinity Large model. Although still smaller than leading models and lagging behind OpenAI and Anthropic on multiple benchmarks, the company plans to develop larger models through a new funding round. NVIDIA has emerged as a key supporter of the U.S. open AI ecosystem, developing its own Nemotron series of models and investing in Reflection AI, Poolside, and Thinking Machines Lab. Industry insiders note that the U.S. open-weight ecosystem remains relatively small, and Chinese models still hold an overall advantage.