Kimi K3 vs Qwen 3.8: China's Open-Weight AI Race

Kimi K3 and Qwen 3.8 Max are the two most important open weight AI models in China's frontier model race. Qwen 3.8 Max now has an official benchmark table, API access, and a planned open weight release, narrowing its evidence gap. Kimi K3, however, still shows stronger independent visibility through various testing and profiles. Qwen is cheaper and more multimodal, while K3 has a clearer record for software engineering and open weight deployment. Kimi K3 arrived first with 2.8 trillion parameters, followed by Alibaba's Qwen 3.8 Max at 2.4 trillion parameters. The bottom line is that developers seeking verified coding performance and available weights might lean towards Kimi K3. Those interested in multimodal agents, long-running workflows, and lower API costs might find Qwen 3.8 Max a more interesting challenger. This matters because it highlights the ongoing competition and advancements in open-weight AI development.