Qwen series comparison
Alibaba's Qwen family: open-source, many sizes and very low pricing, including dedicated code and reasoning models.
All models
Qwen series comparison
Alibaba's Qwen family: open-source, many sizes and very low pricing, including dedicated code and reasoning models.
8 versions (sorted by capability)
| Version | Context | Input | Output | Reasoning | Code | Speed | Open source |
|---|---|---|---|---|---|---|---|
| Qwen2.5-72B | 128K | 开源免费 | 开源免费 | ●●●●○ | ●●●●○ | ●●●●○ | Open |
| Qwen2.5-Coder | 128K | 开源免费 | 开源免费 | ●●●●○ | ●●●●● | ●●●●○ | Open |
| Qwen2.5-Max | 128K | — | — | ●●●●○ | ●●●●○ | ●●●●○ | Closed |
| Qwen3-Coder | — | — | — | ●●●●○ | ●●●●● | ●●●●○ | Open |
| Qwen3-Max | — | — | — | ●●●●● | ●●●●○ | ●●●○○ | Closed |
| Qwen3.5 | — | $0.01 / 1M | — | ●●●●○ | ●●●●○ | ●●●●● | Open |
| Qwen3.6 | — | — | — | ●●●●○ | ●●●●○ | ●●●●● | Open |
| Qwen3.6-Plus | — | — | — | ●●●●○ | ●●●●○ | ●●●●○ | Closed |
Qwen2.5-72BAlibaba's Tongyi Qianwen latest flagship, with the strongest Chinese language capabilities domestically, fully open-source, and supports multimodal.Qwen2.5-CoderAlibaba's specialized code model surpasses Claude 3.5 Sonnet in coding ability, achieving 98.5% on HumanEval, fully open-source.Qwen2.5-MaxTongyi Qianwen 2.5 Max: A large-scale MoE flagship model with comprehensive capabilities comparable to mainstream closed-source models.Qwen3-CoderTongyi Qianwen 3 Code Special: 480B-A35B MoE open source (Apache 2.0), strong code capabilities.Qwen3-MaxTongyi Qianwen 3 Max: A closed-source flagship with over 1T parameters, representing the ceiling of the Tongyi series' capabilities.Qwen3.5Alibaba Tongyi Qianwen 3.5, open-source, fast, and extremely low-cost (starting from approximately $0.01/1M), available in multiple sizes.Qwen3.6Tongyi Qianwen 3.6: 35B-A3B MoE open-source model (Apache 2.0), the latest generation in 2026.Qwen3.6-PlusTongyi Qianwen 3.6 Plus: Closed-source flagship version, released in 2026, with comprehensive capabilities comparable to mainstream closed-source models.