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 | gemini-2.5-flash | 4575.77 t/s Best: 22057.30Worst: 143.99 | 4.39s | 5 | |
| 2 | yuki | 2712.05 t/s Best: 6893.15Worst: 95.27 | 5.02s | 5 | |
| 3 |
Cerebrasapi.cerebras.ai |
| llama3.1-8b |
2142.09 t/s Best: 2534.63Worst: 861.08 |
0.19s |
| 5 |
| 4 | Cerebrasapi.cerebras.ai | gpt-oss-120b | 1920.13 t/s Best: 2415.66Worst: 1587.06 | 0.54s | 5 |
| 5 | Cerebrasapi.cerebras.ai | llama-3.3-70b | 1532.55 t/s Best: 1982.40Worst: 964.64 | 0.25s | 5 |
| 6 | 8 82.157.254.69:300082.157.254.69:3000 | Qwen/Qwen3-32B | 1214.07 t/s Best: 1649.43Worst: 977.23 | 0.91s | 5 |
| 7 | Cerebrasapi.cerebras.ai | qwen-3-235b-a22b-instruct-2507 | 851.89 t/s Best: 1123.67Worst: 610.53 | 12.09s | 10 |
| 8 | Groqapi.groq.com | openai/gpt-oss-20b | 755.20 t/s Best: 955.01Worst: 511.60 | 0.47s | 5 |
| 9 | WONG公益站wzw.pp.ua | grok-4.1 | 729.41 t/s Best: 1814.47Worst: 74.51 | 59.04s | 5 |
| 10 | ChatGTPwww.chatgtp.cn | gpt-4.1-nano-2025-04-14 | 592.40 t/s Best: 1840.57Worst: 213.50 | 1.03s | 10 |
| 11 | 1 156.225.23.250156.225.23.250 | deepseek-ai/DeepSeek-V3.2-Exp | 557.51 t/s Best: 2787.57Worst: 0.00 | 3.46s | 5 |
| 12 | Groqapi.groq.com | openai/gpt-oss-120b | 456.69 t/s Best: 482.07Worst: 427.34 | 0.31s | 10 |
| 13 | New APInew.123nhh.xyz | gemini-flash-lite-latest | 369.22 t/s Best: 551.78Worst: 307.12 | 0.67s | 5 |
| 14 | r realpics.cn:2234realpics.cn:2234 | Qwen3-0.6B-Q8_0.gguf | 340.00 t/s Best: 431.04Worst: 251.73 | 3.08s | 15 |
| 15 | R Realpicsrealpics.cn:2234 | Qwen3-0.6B-Q8_0.gguf | 340.00 t/s Best: 431.04Worst: 251.73 | 3.08s | 15 |
| 16 | r realpics.cn:2234realpics.cn:2234 | Qwen3-0.6B-Q8_0.gguf | 340.00 t/s Best: 431.04Worst: 251.73 | 3.08s | 15 |
| 17 | Groqapi.groq.com | qwen/qwen3-32b | 310.21 t/s Best: 362.78Worst: 275.13 | 0.18s | 5 |
| 18 | r realpics.cn:2234realpics.cn:5002 | Qwen3-0.6B-Q8_0.gguf | 300.87 t/s Best: 305.13Worst: 296.03 | 2.59s | 10 |
| 19 | r realpics.cn:5002realpics.cn:5002 | Qwen3-0.6B-Q8_0.gguf | 300.87 t/s Best: 305.13Worst: 296.03 | 2.59s | 10 |
| 20 | R Realpicsrealpics.cn:5002 | Qwen3-0.6B-Q8_0.gguf | 300.87 t/s Best: 305.13Worst: 296.03 | 2.59s | 10 |
| 21 | 全球AIglobalai.vip | gpt-oss-120b | 284.17 t/s Best: 332.77Worst: 249.78 | 1.50s | 5 |
| 22 | 全球AIglobalai.vip | gemini-2.5-flash-lite | 274.78 t/s Best: 319.23Worst: 248.99 | 0.60s | 5 |
| 23 | e elysiver.h-e.topelysiver.h-e.top | gemini-3-flash-preview | 263.35 t/s Best: 427.47Worst: 188.54 | 6.89s | 5 |
| 24 | integrate.api.nvidia.comintegrate.api.nvidia.com | openai/gpt-oss-120b | 251.35 t/s Best: 282.59Worst: 224.87 | 0.80s | 5 |
| 25 | lansonsamtest.nuiziyyds.com | gemini-2.5-flash | 206.62 t/s Best: 260.40Worst: 142.45 | 7.26s | 5 |
| 26 | p pv4-beta.kxcym.topipv4-beta.kxcym.top:11434 | qwen3-vl:latest | 206.16 t/s Best: 211.21Worst: 200.59 | 6.63s | 5 |
| 27 | i ipv4-beta.kxcym.top:11434ipv4-beta.kxcym.top:11434 | qwen3-vl:latest | 206.16 t/s Best: 211.21Worst: 200.59 | 6.63s | 5 |
| 28 | I Ipv4 Betaipv4-beta.kxcym.top:11434 | qwen3-vl:latest | 206.16 t/s Best: 211.21Worst: 200.59 | 6.63s | 5 |
| 29 | 全球AIglobalai.vip | gemini-2.5-flash | 192.65 t/s Best: 235.51Worst: 144.32 | 9.27s | 5 |
| 30 | r realpics.cn:1234realpics.cn:1234 | gemini-2.5-flash | 188.95 t/s Best: 207.59Worst: 169.05 | 9.25s | 5 |
| 31 | R Realpicsrealpics.cn:1234 | gemini-2.5-flash | 188.95 t/s Best: 207.59Worst: 169.05 | 9.25s | 5 |
| 32 | r realpics.cn:2234realpics.cn:1234 | gemini-2.5-flash | 188.95 t/s Best: 207.59Worst: 169.05 | 9.25s | 5 |
| 33 | 黑与白公益站ai.hybgzs.com | gemini-2.5-flash-lite | 186.31 t/s Best: 355.26Worst: 129.74 | 1.55s | 5 |
| 34 | XJY APIapi.xinjianya.top | gemini-2.5-flash | 183.01 t/s Best: 213.04Worst: 155.85 | 8.28s | 5 |
| 35 | N New APIapi.seosycy.com | gemini-2.0-flash | 181.83 t/s Best: 202.86Worst: 160.24 | 1.79s | 5 |
| 36 | ChatGTPwww.chatgtp.cn | gemini-2.0-flash | 181.28 t/s Best: 199.82Worst: 145.48 | 1.02s | 5 |
| 37 | Z ZenMuxzenmux.ai | z-ai/glm-4.6v-flash | 172.60 t/s Best: 238.26Worst: 118.48 | 9.42s | 5 |
| 38 | o ocool APIone.ocoolai.com | gpt-5.1 | 170.84 t/s Best: 246.21Worst: 119.96 | 1.98s | 5 |
| 39 | a api.linkapi.orgapi.linkapi.org | gemini-2.5-flash-lite-preview-06-17 | 160.85 t/s Best: 213.98Worst: 88.78 | 0.70s | 5 |
| 40 | r realpics.cn:5002realpics.cn:5002 | gpt-oss-20b-MXFP4.gguf | 150.80 t/s Best: 154.26Worst: 147.96 | 2.98s | 10 |
| 41 | R Realpicsrealpics.cn:5002 | gpt-oss-20b-MXFP4.gguf | 150.80 t/s Best: 154.26Worst: 147.96 | 2.98s | 10 |
| 42 | r realpics.cn:2234realpics.cn:5002 | gpt-oss-20b-MXFP4.gguf | 150.80 t/s Best: 154.26Worst: 147.96 | 2.98s | 10 |
| 43 | a ai-hub.square-llm.comai-hub.square-llm.com | google/gemini-3-flash-preview | 150.13 t/s Best: 167.84Worst: 141.36 | 1.45s | 5 |
| 44 | ETOS APIapi.ericterminal.com | moonshotai/kimi-k2-instruct-0905 | 148.30 t/s Best: 165.64Worst: 133.01 | 0.83s | 5 |
| 45 | 智谱AI开放平台open.bigmodel.cn | GLM-4.6V-Flash | 144.43 t/s Best: 193.43Worst: 88.78 | 9.58s | 15 |
| 46 | j jeniya.cnjeniya.cn | gemini-2.5-flash | 139.24 t/s Best: 185.95Worst: 103.00 | 7.45s | 5 |
| 47 | DashScopedashscope.aliyuncs.com | qwen-flash | 134.43 t/s Best: 143.19Worst: 113.93 | 0.50s | 10 |
| 48 | GPTAPI.USwww.gptapi.us | deepseek-v3.1 | 132.30 t/s Best: 154.59Worst: 110.42 | 0.55s | 5 |
| 49 | 简小智API中转站newapi.jianxiaozhi.chat:56897 | deepseek-v3-1-terminus | 126.69 t/s Best: 136.27Worst: 119.57 | 1.03s | 5 |
| 50 | 简 简小智API中转站newapi.jianxiaozhi.chat:56897 | deepseek-v3-1-terminus | 126.69 t/s Best: 136.27Worst: 119.57 | 1.03s | 5 |