In a reasoning test using Arena-Hard, Qwen 2.5-Max achieved 89.4% accuracy, and the result was higher than DeepSeek R1 and when tested on other benchmarks of coding and scientific reasoning, Qwen 2.5 ...
Qwen Max is a powerful AI model with support for 29 languages and can handle 128,000 tokens. Versions of the AI model like Qwen2.5-Coder-32B excel at coding, while QwQ-32B can think and reason. Qwen ...