Leaderboard
Multi-dimensional rankings based on model speed tests and provider health checks. Compare providers, endpoints, and reliability at a glance.
Average tokens generated per second. Higher is better for fast responses.
| Rank | Provider | Model | Throughput | Avg first token latency | Total Tests |
|---|---|---|---|---|---|
| 1 | o3-mini-high | 10126.52 t/s Best: 12035.32Worst: 6606.85 | 3.47s | 5 | |
| 2 | GPT-4o-Mini | 8515.49 t/s Best: 11262.92Worst: 4554.94 | 3.41s | 5 | |
| 3 |
Nahcrof AIai.nahcrof.com |
| deepseek-v3-0324 |
604.52 t/s Best: 2745.61Worst: 79.29 |
8.05s |
| 15 |
| 4 | A AI Toolsplatform.aitools.cfd | deepseek/deepseek-v3 | 548.45 t/s Best: 2288.11Worst: 89.93 | 4.82s | 5 |
| 5 | surtext.pollinations.ai | qwen-reasoning | 400.90 t/s Best: 421.47Worst: 347.00 | 2.67s | 5 |
| 6 | 百 百万APIapi.zhubaiwan.xyz | Groq/qwen-qwq-32b | 382.11 t/s Best: 432.96Worst: 288.56 | 1.36s | 5 |
| 7 | MN APIwww.mnapi.com | QwQ-32B | 297.86 t/s Best: 401.34Worst: 36.73 | 1.09s | 5 |
| 8 | KKSJ-AIapi.kksj.org | gemini-2.0-flash-exp | 231.20 t/s Best: 344.84Worst: 182.85 | 4.36s | 5 |
| 9 | MN APIwww.mnapi.com | gemini-2.0-flash | 220.95 t/s Best: 319.67Worst: 173.39 | 1.30s | 5 |
| 10 | A AI Toolsplatform.aitools.cfd | deepseek/deepseek-v3-0324 | 210.47 t/s Best: 406.45Worst: 31.81 | 4.28s | 5 |
| 11 | 速 速创APIapi.mengzuihuawu.xyz | claude-3-7-sonnet-20250219 | 179.07 t/s Best: 723.79Worst: 30.69 | 9.50s | 5 |
| 12 | MN APIwww.mnapi.com | gemini-2.5-pro-exp-03-25 | 159.02 t/s Best: 255.07Worst: 107.99 | 17.77s | 5 |
| 13 | surtext.pollinations.ai | openai | 152.10 t/s Best: 159.22Worst: 142.24 | 2.77s | 5 |
| 14 | surtext.pollinations.ai | openai-large | 121.85 t/s Best: 150.26Worst: 73.47 | 2.38s | 10 |
| 15 | YUNWU APIyunwu.ai | gemini-2.5-pro-exp-03-25 | 121.26 t/s Best: 164.64Worst: 83.84 | 16.05s | 5 |
| 16 | ChatAnywhereapi.chatanywhere.tech | gpt-3.5-turbo | 121.17 t/s Best: 141.40Worst: 101.17 | 1.41s | 5 |
| 17 | surtext.pollinations.ai | gemini | 119.88 t/s Best: 150.74Worst: 108.77 | 2.53s | 5 |
| 18 | b binaryYukiai.tzpro.xyz | gpt-4o | 114.07 t/s Best: 156.04Worst: 76.49 | 1.11s | 10 |
| 19 | KKSJ-AIapi.kksj.org | gpt-4o-mini | 107.18 t/s Best: 153.56Worst: 84.01 | 2.14s | 5 |
| 20 | 算 算了么 APIapi.suanli.cn | best-free | 102.43 t/s Best: 134.97Worst: 5.30 | 34.49s | 5 |
| 21 | 算 算了么 APIapi.suanli.cn | best-free | 100.39 t/s Best: 126.04Worst: 5.60 | 0.74s | 5 |
| 22 | SiliconFlowapi.siliconflow.cn | Qwen/Qwen2-7B-Instruct | 98.75 t/s Best: 100.45Worst: 95.18 | 0.54s | 5 |
| 23 | 8 8.217.45.96:30028.217.45.96:3002 | gpt-4o | 92.54 t/s Best: 129.53Worst: 34.60 | 1.71s | 10 |
| 24 | KKSJ-AIapi.kksj.org | gpt-4o-2024-11-20 | 88.57 t/s Best: 158.92Worst: 42.12 | 1.52s | 5 |
| 25 | ChatAnywhereapi.chatanywhere.tech | gpt-4o-mini | 86.37 t/s Best: 107.93Worst: 69.76 | 1.64s | 5 |
| 26 | N Newagiaiapi.newagiai.com | gpt-4o-mini | 86.04 t/s Best: 116.56Worst: 48.78 | 1.27s | 5 |
| 27 | 速 速创APIapi.mengzuihuawu.xyz | gpt-4o-2024-11-20 | 84.77 t/s Best: 132.15Worst: 38.25 | 3.24s | 5 |
| 28 | BytesBoostapi.bytesboost.com | gpt-4o | 78.41 t/s Best: 147.24Worst: 47.66 | 1.03s | 5 |
| 29 | A AI Toolsplatform.aitools.cfd | zhipu/glm-4v-flash | 77.34 t/s Best: 96.33Worst: 53.61 | 0.95s | 5 |
| 30 | S Studioapi.studio.nebius.ai | deepseek-ai/DeepSeek-V3-0324-fast | 73.57 t/s Best: 78.88Worst: 63.78 | 0.62s | 5 |
| 31 | 羊羊羊的APIllmapi.koast.top | Qwen/QwQ-32B-Preview | 70.89 t/s Best: 73.33Worst: 67.64 | 0.80s | 5 |
| 32 | SiliconFlowapi.siliconflow.cn | Qwen/QwQ-32B-Preview | 69.75 t/s Best: 73.71Worst: 67.36 | 0.62s | 5 |
| 33 | SiliconFlowapi.siliconflow.cn | deepseek-ai/DeepSeek-R1-Distill-Qwen-14B | 67.68 t/s Best: 81.31Worst: 42.45 | 10.66s | 5 |
| 34 | 箴 箴理科技api.truth-ai.com.cn | jinshu-1.5 | 66.48 t/s Best: 122.84Worst: 27.30 | 1.59s | 10 |
| 35 | SiliconFlowapi.siliconflow.cn | Qwen/Qwen2.5-14B-Instruct | 64.55 t/s Best: 94.18Worst: 34.42 | 0.61s | 5 |
| 36 | 箴 箴理科技api.truth-ai.com.cn | gpt-4o | 63.64 t/s Best: 115.28Worst: 15.34 | 1.81s | 10 |
| 37 | MN APIwww.mnapi.com | claude-3-7-sonnet-20250219 | 62.66 t/s Best: 79.75Worst: 47.90 | 2.52s | 5 |
| 38 | DashScopedashscope.aliyuncs.com | qwen1.5-7b-chat | 59.96 t/s Best: 61.92Worst: 57.43 | 0.89s | 10 |
| 39 | x.aiapi.x.ai | grok-2-1212 | 58.57 t/s Best: 68.84Worst: 51.94 | 0.60s | 10 |
| 40 | 箴 箴理科技api.truth-ai.com.cn | deepseek-ai/DeepSeek-R1-Distill-Qwen-7B | 58.20 t/s Best: 82.47Worst: 31.09 | 12.80s | 10 |
| 41 | SiliconFlowapi.siliconflow.cn | deepseek-ai/DeepSeek-R1-Distill-Qwen-7B | 57.75 t/s Best: 72.45Worst: 40.70 | 12.05s | 5 |
| 42 | 箴 箴理科技api.truth-ai.com.cn | jinshu | 56.96 t/s Best: 99.10Worst: 31.21 | 1.62s | 5 |
| 43 | KKSJ-AIapi.kksj.org | claude-3-7-sonnet-20250219 | 55.56 t/s Best: 72.49Worst: 43.44 | 3.51s | 5 |
| 44 | N Newagiaiapi.newagiai.com | claude-3-7-sonnet-20250219 | 54.37 t/s Best: 63.89Worst: 43.64 | 6.92s | 5 |
| 45 | 箴 箴理科技api.truth-ai.com.cn | qwen-turbo-2024-11-01 | 50.95 t/s Best: 52.56Worst: 46.88 | 0.92s | 5 |
| 46 | N Newagiaiapi.newagiai.com | grok-3 | 49.76 t/s Best: 56.49Worst: 39.24 | 1.90s | 5 |
| 47 | KKSJ-AIapi.kksj.org | gemini-2.0-pro-exp-02-05 | 49.44 t/s Best: 57.33Worst: 42.76 | 1.77s | 5 |
| 48 | New APIapi.lianwusuoai.top | 火山R1-32B | 48.09 t/s Best: 54.28Worst: 31.01 | 22.59s | 5 |
| 49 | 算 算了么 APIapi.suanli.cn | deepseek-r1:7b | 45.86 t/s Best: 71.52Worst: 0.82 | 4.31s | 65 |
| 50 | YUNWU APIyunwu.ai | qwq-plus-latest | 43.91 t/s Best: 49.41Worst: 36.58 | 19.49s | 5 |