synopsis

Qwen3 AI models are "hybrid", in that they can take time to "reason" through complex problems or quickly answer simpler requests.

Chinese tech heavyweight Alibaba (BABA) on Tuesday released Qwen3, a new set of artificial intelligence models it claims match up to and, in some cases, outperform those from OpenAI and Google (GOOGL).

Alibaba said its latest release combines traditional AI capabilities with sophisticated dynamic reasoning, offering what the company describes as a more flexible and efficient platform for app and software developers.

Alibaba has been releasing new AI models at a breakneck pace. Combined with DeepSeek, which stunned the world by developing sophisticated AI at a fraction of the cost of American models, these developments have made the West take notice of China's AI prowess.

Alibaba's U.S.-listed shares rose a whopping 75% this year till mid-February before crashing with other Chinese stocks as threats from U.S. President Donald Trump's trade policies became imminent.

Chinese search leader Baidu (BIDU) also upped the stakes with the release of its Ernie 4.5 Turbo and reasoning-focused Ernie X1 Turbo models last Friday.

Alibaba said Qwen3 models are "hybrid", in that they can take time to "reason" through complex problems or quickly answer simpler requests. Reasoning enables the models to effectively fact-check themselves, similar to capabilities in OpenAI's o3.

"We have seamlessly integrated thinking and non-thinking modes, offering users the flexibility to control the thinking budget," the Qwen team said in a blog post published on Tuesday.

"This design enables users to configure task-specific budgets with greater ease."

Most Qwen3 models can be downloaded under an open license from developer platforms like Hugging Face and GitHub.

“Qwen3” was trending on Google Search in the U.S. at the time of writing.

For updates and corrections, email newsroom[at]stocktwits[dot]com.<