Claude Fable 5.2 scored 65, putting it 22 points ahead of Alibaba's flagship model.
This story is only covered by news sources that have yet to be evaluated by the independent media monitoring agencies we use to assess the quality and reliability of news outlets on our platform. Learn more here.
Alibaba unveiled its new flagship model, the Qwen4-Max, on Monday, November 16, and the company's own benchmarks compare it to Anthropic's Claude Opus 5.5…
To enrich his new AI model without weighing up the calculations in the same proportions, Alibaba added an extensive table of digital representations. A cost-effective approach to calculation, but which requires space to store weights.
Alibaba has introduced Qwen4-Max. The AI model includes 2.4 trillion parameters and activates 71 billion per token. In the internal Work360 test, the average cost per task is to be reduced by around 60% compared to Qwen3.8-Max.