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-lite-preview-06-17 | 228523.98 t/s Best: 275494.22Worst: 167422.60 | 4.79s | 5 | |
| 2 | gemini-2.5-pro | 226468.01 t/s Best: 266694.80Worst: 184690.62 | 29.88s | 5 |
| 3 | ChatAnywhereapi.chatanywhere.org | gpt-5-mini | 155739.72 t/s Best: 241216.90Worst: 72339.91 | 12.30s | 5 |
| 4 | C ClawCloud Runjhgptmycidwg.eu-central-1.clawcloudrun.com | gemini-2.5-flash-lite-search | 115243.74 t/s Best: 159414.00Worst: 77574.79 | 8.28s | 5 |
| 5 | New APIapi.huajinet.link | gemini-2.5-flash | 17911.37 t/s Best: 28154.49Worst: 5924.55 | 14.53s | 5 |
| 6 | 专 专盾Procdnapi2.dashi.party | gemini-2.5-pro-search | 12441.96 t/s Best: 32439.33Worst: 0.00 | 16.06s | 5 |
| 7 | 专 专盾Procdnapi2.dashi.party | gemini-2.5-pro-non-thinking | 10648.97 t/s Best: 29606.26Worst: 0.00 | 10.79s | 10 |
| 8 | 专 专盾Procdnapi2.dashi.party | gemini-2.5-flash-lite | 5461.52 t/s Best: 8519.16Worst: 1500.97 | 3.24s | 5 |
| 9 | surtext.pollinations.ai | evil | 1475.97 t/s Best: 2603.89Worst: 43.96 | 0.81s | 25 |
| 10 | Cotton APIgemini.nkbpal.cn | gpt-oss-120b | 771.04 t/s Best: 1060.54Worst: 537.86 | 0.96s | 5 |
| 11 | 专 专盾Procdnapi2.dashi.party | qwen-3-32b | 571.61 t/s Best: 721.99Worst: 420.20 | 1.10s | 5 |
| 12 | S SWT-APIapi.lhyb.dpdns.org | qwen-3-32b-turbo | 515.39 t/s Best: 574.61Worst: 367.72 | 2.33s | 5 |
| 13 | surtext.pollinations.ai | qwen | 514.50 t/s Best: 1204.31Worst: 50.42 | 1.89s | 5 |
| 14 | Yuegleapi.yuegle.com | gemini-2.5-flash-lite-preview-06-17 | 401.84 t/s Best: 448.51Worst: 330.74 | 0.61s | 5 |
| 15 | A AI Proxy Serviceai-proxy.4ba-cn.co | google/gemini-2.5-flash-lite | 363.58 t/s Best: 410.21Worst: 288.92 | 1.42s | 5 |
| 16 | New APIoneapi.352287.xyz | allam-2-7b | 337.15 t/s Best: 517.78Worst: 19.95 | 0.27s | 5 |
| 17 | Yuegleapi.yuegle.com | gemini-2.5-flash-lite | 335.31 t/s Best: 358.52Worst: 315.00 | 0.52s | 5 |
| 18 | Undy APIvip.undyingapi.com | gemini-2.5-flash-lite | 296.41 t/s Best: 355.10Worst: 175.06 | 1.49s | 5 |
| 19 | S SWT-APIapi.lhyb.dpdns.org | qwen-3-235b-a22b-instruct-2507 | 271.04 t/s Best: 361.74Worst: 127.90 | 1.09s | 10 |
| 20 | A AI Toolsplatform.aitools.cfd | deepseek/deepseek-v3-0324 | 263.53 t/s Best: 3035.26Worst: 18.19 | 5.14s | 95 |
| 21 | F Freddy Greveai-api.freddygreve.com | p-openai-fast | 254.97 t/s Best: 281.16Worst: 228.06 | 1.55s | 5 |
| 22 | surtext.pollinations.ai | openai-roblox | 248.43 t/s Best: 280.07Worst: 235.96 | 3.76s | 5 |
| 23 | Undy APIvip.undyingapi.com | gpt-5-nano | 247.78 t/s Best: 258.42Worst: 232.59 | 6.22s | 5 |
| 24 | Hugging Facerouter.huggingface.co | openai/gpt-oss-120b:novita | 240.32 t/s Best: 265.38Worst: 202.19 | 1.26s | 5 |
| 25 | Hugging Facerouter.huggingface.co | openai/gpt-oss-120b:novita | 240.32 t/s Best: 265.38Worst: 202.19 | 1.26s | 5 |
| 26 | integrate.api.nvidia.comintegrate.api.nvidia.com | openai/gpt-oss-20b | 239.61 t/s Best: 250.77Worst: 211.54 | 10.88s | 5 |
| 27 | J Just a moment...api.qwq.chat | gpt-oss-120b | 236.16 t/s Best: 248.43Worst: 199.06 | 1.55s | 5 |
| 28 | 拼好站new.xigua.wiki | o4-mini | 231.62 t/s Best: 471.98Worst: 115.12 | 9.23s | 5 |
| 29 | 8 8.138.108.72:30008.138.108.72:3000 | openai/gpt-oss-120b | 231.52 t/s Best: 241.58Worst: 203.33 | 1.87s | 5 |
| 30 | Mineai.081007.xyz | gpt-oss-20b | 229.65 t/s Best: 278.43Worst: 156.94 | 1.77s | 10 |
| 31 | 鲨 鲨鱼魔法openai.sharkmagic.com.cn | [503]gemini-flash-lite | 229.64 t/s Best: 308.20Worst: 158.99 | 2.17s | 5 |
| 32 | F Freddy Greveai-api.freddygreve.com | p-gemini | 228.63 t/s Best: 339.95Worst: 140.29 | 2.82s | 5 |
| 33 | Hugging Facerouter.huggingface.co | openai/gpt-oss-120b | 225.67 t/s Best: 241.37Worst: 198.53 | 0.92s | 5 |
| 34 | Hugging Facerouter.huggingface.co | openai/gpt-oss-120b | 225.67 t/s Best: 241.37Worst: 198.53 | 0.92s | 5 |
| 35 | surtext.pollinations.ai | openai | 225.32 t/s Best: 277.36Worst: 192.03 | 2.40s | 5 |
| 36 | F Freddy Greveai-api.freddygreve.com | p-openai | 223.10 t/s Best: 241.57Worst: 186.17 | 1.54s | 5 |
| 37 | New APIoneapi.352287.xyz | gemini-2.5-flash-lite-ts | 214.83 t/s Best: 293.14Worst: 158.86 | 1.17s | 5 |
| 38 | 专 专盾Procdnapi2.dashi.party | Qwen/Qwen3-235B-A22B | 209.58 t/s Best: 707.69Worst: 40.33 | 1.62s | 15 |
| 39 | Mineai.081007.xyz | gemini-2.5-flash | 208.59 t/s Best: 276.67Worst: 166.48 | 8.73s | 5 |
| 40 | 1 180.76.61.29:3000180.76.61.29:3000 | gemini-2.5-flash | 207.67 t/s Best: 455.94Worst: 96.71 | 6.78s | 5 |
| 41 | 鲨 鲨鱼魔法openai.sharkmagic.com.cn | [503]gemini-flash | 199.70 t/s Best: 241.45Worst: 180.76 | 10.04s | 5 |
| 42 | J Just a moment...api.qwq.chat | (gold)gemini-2.5-flash | 192.64 t/s Best: 208.66Worst: 167.20 | 9.01s | 5 |
| 43 | Yuegleapi.yuegle.com | gemini-2.5-flash-cheap | 187.61 t/s Best: 234.57Worst: 149.06 | 8.15s | 5 |
| 44 | 1 139.196.127.92:3002139.196.127.92:3002 | models/gemini-2.5-flash | 184.25 t/s Best: 203.81Worst: 159.14 | 8.26s | 5 |
| 45 | Cotton APIgemini.nkbpal.cn | gemini-2.5-flash-nothinking | 174.04 t/s Best: 198.57Worst: 143.55 | 0.82s | 5 |
| 46 | YUNWU APIyunwu.ai | gemini-2.5-pro-thinking | 165.57 t/s Best: 405.64Worst: 90.66 | 4.10s | 5 |
| 47 | 共绩算力d08011731-minicpm4-8blatest-2824-9z9f7zk2-11434.550c.cloud | minicpm4-8b:latest | 159.84 t/s Best: 170.29Worst: 153.79 | 0.96s | 5 |
| 48 | d d08011731-minicpm4-8blatest-2824-9z9f7zk2-11434.550c.cloudd08011731-minicpm4-8blatest-2824-9z9f7zk2-11434.550c.cloud | minicpm4-8b:latest | 159.84 t/s Best: 170.29Worst: 153.79 | 0.96s | 5 |
| 49 | Hugging Facerouter.huggingface.co | openai/gpt-oss-20b:novita | 155.67 t/s Best: 191.01Worst: 132.06 | 3.56s | 5 |
| 50 | Hugging Facerouter.huggingface.co | openai/gpt-oss-20b:novita | 155.67 t/s Best: 191.01Worst: 132.06 | 3.56s | 5 |