Alibaba Cloud’s newest AI mannequin, Qwen3-Max-Pondering, is staking a declare as one of many world’s most superior reasoning engines after posting benchmark outcomes that delivered aggressive outcomes towards main fashions from Google and OpenAI.
In a weblog submit, Alibaba stated the mannequin was educated utilizing expanded capability and large-scale computing assets, together with reinforcement studying, which led to enhancements in factual accuracy, reasoning, instruction following, alignment with human preferences, and agent-style capabilities.
“On 19 established benchmarks, it demonstrates efficiency similar to main fashions resembling GPT-5.2-Pondering, Claude-Opus-4.5, and Gemini 3 Professional,” the corporate stated.
