AIThe Registerabout 3 hours ago
China's open model blitz pressures US rivals
9 min read
Alibaba launched Qwen 3.8-Max, a 2.4 trillion-parameter open-weight model matching Anthropic and OpenAI performance, with weights releasing next week. DeepSeek V4-Flash-0731, a 284-billion-parameter model, outperforms DeepSeek V4 Pro by 14% and costs 40% less per task than OpenAI's GPT-5.6 Luna.
Level
Hype check
Tap to vote and see what everyone thinks.
