智谱AI开放平台open.bigmodel.cn
Leaderboard
Multi-dimensional rankings based on model speed tests and provider health checks. Compare providers, endpoints, and reliability at a glance.
Average time to first token. Lower is better for responsiveness.
| Rank | Provider | Model | First Token Latency | Avg tokens per second | Total Tests |
|---|---|---|---|---|---|
| 1 | glm-4-flash-250414 | 0.31 s Best: 0.18Worst: 0.65 | 58.60t/s | 10 | |
| 2 | DeepSeek-v3 | 0.38 s Best: 0.26Worst: 0.74 | 41.60t/s | 10 |
| 3 | qwen3:0.6b | 0.40 s Best: 0.39Worst: 0.40 | 239.34t/s | 5 |
| 4 | qwen3:0.6b | 0.40 s Best: 0.39Worst: 0.40 | 239.34t/s | 5 |
| 5 | deepseek_7b | 0.45 s Best: 0.39Worst: 0.47 | 4773.40t/s | 5 |
| 6 | gpt-4o-2024-05-13 | 0.49 s Best: 0.43Worst: 0.56 | 100.94t/s | 5 |
| 7 | Qwen2.5-32B-Instruct | 0.51 s Best: 0.49Worst: 0.52 | 37.44t/s | 5 |
| 8 | Qwen2.5-32B-Instruct | 0.51 s Best: 0.49Worst: 0.52 | 37.44t/s | 5 |
| 9 | qwen3:30b-a3b-q8_0 | 0.55 s Best: 0.54Worst: 0.56 | 84.32t/s | 5 |
| 10 | qwen3:30b-a3b-q8_0 | 0.55 s Best: 0.54Worst: 0.56 | 84.32t/s | 5 |
| 11 | deepseek-r1:32b | 0.56 s Best: 0.47Worst: 0.84 | 39.94t/s | 5 |
| 12 | deepseek-r1:32b | 0.56 s Best: 0.47Worst: 0.84 | 39.94t/s | 5 |
| 13 | /root/models/Qwen/Qwen3-4B | 0.56 s Best: 0.35Worst: 1.34 | 72.15t/s | 5 |
| 14 | /root/models/Qwen/Qwen3-4B | 0.56 s Best: 0.35Worst: 1.34 | 72.15t/s | 5 |
| 15 | qwen3:30b | 0.64 s Best: 0.41Worst: 1.90 | 38.64t/s | 30 |
| 16 | inception/mercury-coder-small-beta | 0.72 s Best: 0.53Worst: 1.00 | 331.42t/s | 5 |
| 17 | qwen/qwen2.5-vl-72b | 0.72 s Best: -Worst: 1.88 | 0.00t/s | 10 |
| 18 | grok-3-fast-beta | 0.75 s Best: 0.58Worst: 1.36 | 89.18t/s | 5 |
| 19 | qwen3:30b | 0.77 s Best: 0.52Worst: 1.35 | 61.59t/s | 10 |
| 20 | fradser/deeptranslate-r2-4b:latest | 0.78 s Best: 0.41Worst: 3.96 | 123.70t/s | 10 |
| 21 | fradser/deeptranslate-r2-4b:latest | 0.78 s Best: 0.41Worst: 3.96 | 123.70t/s | 10 |
| 22 | fradser/deeptranslate-r2-4b:latest | 0.78 s Best: 0.41Worst: 3.96 | 123.70t/s | 10 |
| 23 | qwen/qwen2.5-72b | 0.80 s Best: 0.65Worst: 1.11 | 0.00t/s | 5 |
| 24 | gemini-2.0-flash-lite-preview-02-05 | 0.81 s Best: 0.76Worst: 0.87 | 146.97t/s | 5 |
| 25 | gpt-4.1-nano-2025-04-14 | 0.82 s Best: 0.57Worst: 1.13 | 152.16t/s | 5 |
| 26 | zhipu/glm-z1-32b | 0.83 s Best: 0.73Worst: 1.06 | 0.00t/s | 5 |
| 27 | zhipu/glm-4-flash | 0.84 s Best: 0.36Worst: 2.86 | 39.23t/s | 100 |
| 28 | doubao-lite-4k-character-240828 | 0.84 s Best: 0.51Worst: 1.31 | 52.00t/s | 5 |
| 29 | meta/llama-4-scout | 0.93 s Best: 0.76Worst: 1.17 | 0.00t/s | 5 |
| 30 | meta/llama-4-maverick | 0.93 s Best: 0.70Worst: 1.30 | 0.00t/s | 5 |
| 31 | QwQ-32B | 0.94 s Best: 0.75Worst: 1.12 | 47.96t/s | 10 |
| 32 | qwen-max-0428 | 0.97 s Best: 0.70Worst: 1.89 | 16.86t/s | 5 |
| 33 | qwen-max | 0.98 s Best: 0.68Worst: 2.17 | 24.47t/s | 10 |
| 34 | meta-llama/Llama-4-Scout-17B-16E-Instruct | 1.04 s Best: 0.75Worst: 1.47 | 118.72t/s | 5 |
| 35 | hunyuan-lite | 1.04 s Best: 0.93Worst: 1.31 | 133.10t/s | 5 |
| 36 | deepseek-ai/DeepSeek-Prover-V2-671B | 1.05 s Best: 0.80Worst: 1.68 | 80.53t/s | 10 |
| 37 | deepseek-ai/DeepSeek-V3-0324 | 1.08 s Best: 0.78Worst: 1.50 | 74.08t/s | 5 |
| 38 | zhipu/glm-4v-flash | 1.11 s Best: 0.82Worst: 2.08 | 57.09t/s | 5 |
| 39 | DeepSeek-V3 | 1.19 s Best: 1.05Worst: 1.67 | 39.21t/s | 5 |
| 40 | deepseek-v3-250324 | 1.20 s Best: 0.77Worst: 1.63 | 25.68t/s | 5 |
| 41 | opengvlab/internvl3-2b:free | 1.22 s Best: 1.05Worst: 1.39 | 379.88t/s | 5 |
| 42 | google/gemma-3-27b | 1.23 s Best: 1.14Worst: 1.34 | 0.00t/s | 10 |
| 43 | qwen3:30b-a3b | 1.27 s Best: 0.50Worst: 4.30 | 116.49t/s | 5 |
| 44 | qwen3:30b-a3b | 1.27 s Best: 0.50Worst: 4.30 | 116.49t/s | 5 |
| 45 | qwen-max-latest | 1.32 s Best: 1.00Worst: 2.18 | 19.27t/s | 5 |
| 46 | zhipu/glm-4-32b | 1.37 s Best: 1.04Worst: 2.17 | 0.00t/s | 10 |
| 47 | google/gemini-2.0-flash-exp | 1.49 s Best: 1.28Worst: 1.95 | 0.00t/s | 5 |
| 48 | unsloth/qwen3:14b-q8_0 | 1.51 s Best: 0.50Worst: 5.49 | 61.79t/s | 5 |
| 49 | unsloth/qwen3:14b-q8_0 | 1.51 s Best: 0.50Worst: 5.49 | 61.79t/s | 5 |
| 50 | deepseek-v3-0324 | 1.54 s Best: 1.33Worst: 2.08 | 20.44t/s | 5 |
