Qwen3
Think Deeper or Act Faster
120 followers
Qwen3 is the large language model series developed by Qwen team, Alibaba Cloud. - QwenLM/Qwen3
Subscribe
Sign in
120 followers
Qwen3 is the large language model series developed by Qwen team, Alibaba Cloud. - QwenLM/Qwen3
Hi everyone!
Qwen3 is here! It's the latest family of open-weight large language models just released by the Alibaba Qwen team. This is a significant drop, including six dense models (0.6B to 32B) and two MoE models (30B & 235B).
A really interesting feature across these models is the Hybrid Thinking Mode. You can choose – let the model respond quickly, or activate a deeper, step-by-step reasoning process before it answers, giving you flexibility between speed and thoroughness.
Performance looks very competitive. The flagship 235B MoE is positioned against top models like DeepSeek-R1 and o3-mini, while even the smaller dense models show strong results, with the 4B apparently rivaling their previous 72B Instruct model.
They've focused on improving coding, math, and agent capabilities across the board.
You can try them directly in Qwen Chat (web and app) or run them locally via tools like Ollama.
Elisi : AI-powered Goal Management App
Thanks for the heads-up — this drop sounds huge! 🚀
Hybrid Thinking Mode especially caught my attention — being able to switch between speed and deeper reasoning on demand feels like it could be a real game-changer for different workflows (quick chats vs. complex problem solving). Also impressive that even the 4B dense model is showing performance close to much larger previous gens.
Definitely going to try Qwen3 soon — curious to see how it compares to DeepSeek and o3-mini in real-world coding and agent tasks. 🙌
Are you planning to benchmark it locally too? Would love to hear your impressions if you do!
Haye
Besides Alibaba Cloud, does Qwen have its own API channels? Every time I want to use it, opening Alibaba Cloud just makes me feel discouraged.