Alibaba announced the release and open-sourcing of its latest large language model, Qwen3. This release includes two Mixture-of-Experts (MoE) models and six dense models, with parameter sizes ranging from 0.6B to 235B.
1 / 2
Comments
You can experience Qwen3 through the Qwen chat web version and APP. https://chat.qwen.ai/
Dual Inference Modes: Qwen3 supports both a 'Thinking Mode' (suitable for complex problems requiring step-by-step reasoning) and a 'Non-Thinking Mode' (for quick responses to simple questions).
Blog: https://qwenlm.github.io/blog/qwen3/