DashScope提供AI模型和API，主要用于自然语言处理和计算机视觉应用。
| 模型 | 速度 | 延迟 | 测试数 |
|---|---|---|---|
| qwen-mt-turbo | 5403.35 t/s | 1.02s | 5 |
| qwen3-0.6b | 252.65 t/s | 2.49s | 10 |
| qwen3-1.7b | 155.43 t/s | 4.85s | 10 |
| qwen3-next-80b-a3b-instruct | 153.69 t/s | 3.99s | 5 |
| qwen3-next-80b-a3b-instruct | 153.69 t/s | 3.99s | 5 |
| qwen3-vl-flash | 147.58 t/s | 1.00s | 5 |
| qwen3-vl-flash | 147.58 t/s | 1.00s | 5 |
| deepseek-r1-distill-qwen-1.5b | 135.25 t/s | 6.13s | 5 |
| qwen-flash | 128.53 t/s | 0.63s | 20 |
| qwen-vl-plus-2025-07-10 | 112.44 t/s | 0.56s | 65 |
| qwen2.5-7b-instruct-1m | 92.57 t/s | 0.93s | 10 |
| qwen2.5-7b-instruct-1m | 92.57 t/s | 0.93s | 10 |
| deepseek-r1-distill-qwen-7b | 92.17 t/s | 9.98s | 5 |
| qwen-flash-2025-07-28 | 91.82 t/s | 1.48s | 5 |
| qwen3-8b | 91.25 t/s | 8.67s | 10 |
| tongyi-intent-detect-v3 | 87.54 t/s | 0.72s | 10 |
| qwen-vl-plus | 69.83 t/s | 1.00s | 15 |
| tongyi-xiaomi-analysis-pro | 65.76 t/s | 0.64s | 5 |
| qwen2.5-coder-0.5b-instruct | 60.38 t/s | 0.75s | 5 |
| qwen2.5-coder-0.5b-instruct | 60.38 t/s | 0.75s | 5 |
| 时间 | 模型 | 速度 | 延迟 |
|---|---|---|---|
| Jan 12, 07:10 PM | qwen3-235b-a22b-thinking-2507 | 68.48 t/s | 8.90s |
| Jan 12, 07:09 PM | tongyi-xiaomi-analysis-pro | 65.76 t/s | 0.64s |
| Jan 12, 07:01 PM | qwen3-max-2025-09-23 | 28.32 t/s | 1.36s |
| Jan 12, 06:57 PM | qwen-max-latest | 37.07 t/s | 0.79s |
| Jan 12, 06:35 PM | deepseek-r1 | 37.34 t/s | 13.81s |
| Jan 12, 06:25 PM | qwen-plus-2025-12-01 | 56.15 t/s | 0.67s |
| Jan 12, 06:24 PM | qwen-plus-latest | 57.95 t/s | 0.69s |
| Jan 12, 06:20 PM | kimi-k2-thinking | 43.87 t/s | 14.94s |
| Jan 12, 06:18 PM | deepseek-v3.2 | 32.84 t/s | 0.73s |
| Jan 11, 01:36 AM | deepseek-v3.2 | 35.95 t/s | 0.66s |