Qwen 3.6 (通义千问)
FREEMIUMAlibaba's flagship open-weight model series. Qwen3.6 Plus scores 78.8% on SWE-bench Verified. Qwen3.6-27B achieves 77.2% as a dense model. Top open-weight coding model with Apache 2.0 license.
📊 Ratings
4.3/ 5
4.3
Features
4.3
Ease of Use
4.3
Value
4.3
Support
✅ Pros
- ✓Best open-weight coding model
- ✓Strong competitive programming (83.6% LiveCodeBench)
- ✓Extremely affordable at $0.50/$2 per M tokens
- ✓Multiple sizes available (27B, 35B-A3B MoE)
- ✓Fully open-source with Apache 2.0 license
❌ Cons
- ✗Trails Claude and GPT on SWE-bench Pro
- ✗Limited English documentation
- ✗Smaller ecosystem than US competitors
Comments (0)
Join the conversation
Log in to commentNo comments yet. Be the first to share your thoughts!