Measured over the last 30 days
- Models Tracked
- 100
- Avg Tokens / Second
- 17.70
- Avg Time to First Token (ms)
- 10249.29
- Last Updated
- Sep 1, 2026
All novita models
100 tracked · fastest first| Model | Avg Toks/Sec | Min | Max | Avg TTF (ms) |
|---|---|---|---|---|
| Llama 3.1 8B Instruct | 53.60 | 41.50 | 61.20 | 830.00 |
| Llama 3 8B Lunaris | 48.90 | 44.30 | 51.30 | 640.00 |
| Llama 4 Maverick | 48.80 | 38.20 | 58.90 | 660.00 |
| Nemotron 3 Nano 30B A3B | 47.10 | 32.90 | 66.20 | 1220.00 |
| Ling-3.0-flash | 45.10 | 41.00 | 47.00 | 1280.00 |
| Qwen3 Next 80B A3B Instruct | 44.80 | 24.60 | 55.60 | 1010.00 |
| Qwen3 Coder 30B A3B Instruct | 43.60 | 25.40 | 63.80 | 1100.00 |
| Qwen3 Coder 30B A3B Instruct | 42.00 | 41.20 | 42.80 | 1040.00 |
| Qwen3 VL 30B A3B Instruct | 40.20 | 32.30 | 51.40 | 1160.00 |
| Ling-2.6-flash | 40.10 | 21.20 | 57.50 | 910.00 |
| Ling-3.0-flash | 39.80 | 35.30 | 43.30 | 1460.00 |
| Ling-2.6-flash | 39.00 | 34.70 | 42.20 | 900.00 |
| Ling-2.6-1T | 36.00 | 34.10 | 37.30 | 1090.00 |
| Qwen3 Coder Next | 33.30 | 26.70 | 41.80 | 1620.00 |
| Llama 4 Maverick | 32.40 | 21.00 | 38.80 | 1350.00 |
| Kimi K2 0711 | 31.70 | 31.70 | 31.70 | 720.00 |
| Llama 3.1 Euryale 70B v2.2 | 31.00 | 19.60 | 38.20 | 650.00 |
| Qwen3 Coder 480B A35B | 30.30 | 30.30 | 30.30 | 1200.00 |
| Llama 4 Scout | 29.40 | 22.50 | 43.10 | 780.00 |
| Ring-2.6-1T | 29.20 | 29.20 | 29.20 | 1490.00 |
| Ling-2.6-1T | 27.90 | 5.92 | 42.00 | 1580.00 |
| Llama 3.1 8B Instruct | 27.80 | 27.80 | 27.80 | 2010.00 |
| MiniMax M3 | 27.30 | 22.00 | 32.40 | 2120.00 |
| Kimi K2 0905 | 27.10 | 27.10 | 27.10 | 800.00 |
| Gemma 3 27B | 25.00 | 23.60 | 26.50 | 930.00 |
| Kimi K2 0905 | 25.00 | 22.50 | 27.50 | 1090.00 |
| Qwen3 235B A22B Instruct 2507 | 24.70 | 20.60 | 29.90 | 990.00 |
| Llama 3.3 70B Instruct | 23.60 | 14.90 | 32.50 | 920.00 |
| Kimi K2 0711 | 23.60 | 22.50 | 25.80 | 1220.00 |
| Ring-2.6-1T | 23.30 | 20.20 | 26.00 | 2250.00 |
| DeepSeek V3 | 21.50 | 20.80 | 22.50 | 1320.00 |
| Gemma 3 27B | 21.40 | 13.40 | 28.90 | 1130.00 |
| Qwen3 VL 235B A22B Instruct | 21.30 | 12.90 | 35.00 | 780.00 |
| DeepSeek V3.2 | 21.20 | 12.50 | 26.10 | 1560.00 |
| Qwen3 VL 235B A22B Instruct | 21.00 | 8.73 | 33.30 | 1010.00 |
| DeepSeek V3 0324 | 20.90 | 17.90 | 22.70 | 1560.00 |
| gpt-oss-120b | 19.60 | 14.30 | 27.90 | 2740.00 |
| DeepSeek V3 0324 | 19.40 | 16.90 | 23.00 | 1620.00 |
| ERNIE 4.5 VL 424B A47B | 18.80 | 15.00 | 21.80 | 1730.00 |
| DeepSeek V3 | 18.40 | 15.00 | 22.80 | 1810.00 |
| Gemma 4 26B A4B | 18.10 | 4.30 | 27.70 | 1160.00 |
| DeepSeek V3.2 Exp | 17.40 | 16.20 | 18.50 | 1510.00 |
| DeepSeek V3.1 Terminus | 17.10 | 15.20 | 18.60 | 1770.00 |
| DeepSeek V3.2 Exp | 17.10 | 14.00 | 21.80 | 1700.00 |
| Qwen3 235B A22B Instruct 2507 | 17.00 | 8.92 | 25.10 | 2800.00 |
| DeepSeek V3.1 | 16.20 | 15.00 | 18.30 | 1700.00 |
| DeepSeek V3.1 Terminus | 16.20 | 15.00 | 17.80 | 1960.00 |
| ERNIE 4.5 VL 424B A47B | 15.60 | 6.28 | 21.10 | 1930.00 |
| DeepSeek V3.1 | 15.50 | 12.80 | 17.00 | 2010.00 |
| GLM 5.2 | 14.90 | 9.31 | 20.80 | 3380.00 |
| GLM 5.2 | 14.40 | 11.70 | 16.80 | 3300.00 |
| Auto Router (Beta) | 14.30 | 14.30 | 14.30 | 3990.00 |
| gpt-oss-20b | 14.00 | 12.10 | 15.90 | 3940.00 |
| Llama 3.3 70B Instruct | 13.20 | 5.94 | 21.30 | 3210.00 |
| DeepSeek V4 Flash 0731 | 13.10 | 8.07 | 17.10 | 4610.00 |
| MiMo-V2.5-Pro | 12.60 | 7.85 | 15.50 | 4350.00 |
| MiniMax M1 | 12.50 | 10.30 | 14.40 | 6440.00 |
| GLM 4.5 Air | 11.60 | 9.69 | 13.50 | 4910.00 |
| Gemma 4 31B | 11.10 | 11.10 | 11.10 | 1400.00 |
| WizardLM-2 8x22B | 10.70 | 10.40 | 11.00 | 1180.00 |
| WizardLM-2 8x22B | 10.40 | 5.57 | 19.30 | 2810.00 |
| DeepSeek V4 Flash 0423 | 9.70 | 2.59 | 14.60 | 9630.00 |
| Qwen3.8 2.4T A95B | 9.16 | 9.16 | 9.16 | 5520.00 |
| GLM 5.1 | 8.56 | 1.42 | 21.90 | 24400.00 |
| GLM 4.5 Air | 8.52 | 7.90 | 8.94 | 6500.00 |
| MiMo-V2.5 | 8.24 | 3.60 | 13.30 | 9410.00 |
| MiniMax M2.1 | 8.03 | 6.65 | 9.68 | 8690.00 |
| GLM 5 | 7.97 | 1.97 | 24.60 | 20300.00 |
| Mistral Nemo | 7.86 | 7.86 | 7.86 | 5120.00 |
| MiniMax M2 | 7.15 | 5.09 | 7.95 | 9240.00 |
| MiniMax M2.1 | 6.02 | 3.49 | 11.70 | 14600.00 |
| Kimi K2.7 Code | 5.84 | 1.72 | 9.98 | 15800.00 |
| Kimi K2.5 | 5.76 | 5.76 | 5.76 | 9610.00 |
| GLM 4.5V | 5.49 | 4.03 | 6.44 | 11500.00 |
| MiniMax M2.7 | 5.40 | 2.21 | 22.50 | 14600.00 |
| MiniMax M2.5 | 5.37 | 4.01 | 6.81 | 13600.00 |
| Kimi K2 Thinking | 5.31 | 5.31 | 5.31 | 10500.00 |
| Qwen3 VL 235B A22B Thinking | 5.29 | 4.10 | 7.01 | 11400.00 |
| Kimi K2.5 | 5.22 | 4.75 | 5.93 | 10700.00 |
| Kimi K2 Thinking | 4.73 | 4.00 | 5.29 | 12200.00 |
| Qwen3 235B A22B Thinking 2507 | 4.66 | 2.72 | 8.35 | 16200.00 |
| GLM 4.5V | 4.65 | 3.47 | 5.25 | 13600.00 |
| Qwen3 VL 235B A22B Thinking | 4.41 | 2.03 | 7.24 | 14700.00 |
| Qwen3.5-122B-A10B | 3.90 | 2.14 | 5.35 | 18500.00 |
| Qwen3.5-27B | 3.27 | 3.05 | 3.53 | 18900.00 |
| Qwen3.5 397B A17B | 3.22 | 2.50 | 3.92 | 19700.00 |
| R1 Distill Llama 70B | 3.11 | 1.51 | 4.80 | 22000.00 |
| R1 Distill Llama 70B | 3.07 | 2.89 | 3.25 | 18000.00 |
| GLM 4.6 | 3.05 | 2.36 | 3.76 | 20700.00 |
| Kimi K2.6 | 2.90 | 1.82 | 4.52 | 24100.00 |
| R1 | 2.84 | 1.24 | 4.64 | 25100.00 |
| GLM 4.7 Flash | 2.73 | 2.73 | 2.73 | 22400.00 |
| R1 | 2.70 | 2.14 | 3.26 | 21300.00 |
| Hy3 | 2.66 | 2.28 | 3.31 | 24200.00 |
| GLM 4.7 | 2.55 | 1.64 | 3.30 | 26500.00 |
| MiniMax M2.7 | 2.32 | 1.31 | 4.10 | 37500.00 |
| GLM 4.6V | 1.49 | 0.85 | 2.66 | 46100.00 |
| Qwen3.8 27B | 1.15 | 0.65 | 1.75 | 65900.00 |
| GLM 5.3 | 1.14 | 0.10 | 2.18 | 23800.00 |
| GLM 4.6V | 0.95 | 0.95 | 0.95 | 65400.00 |
Frequently Asked Questions
Which novita model is fastest?
Based on recent tests, Llama 3.1 8B Instruct shows the highest average throughput among tracked novita models.
How many recent measurements feed this dashboard?
This provider summary aggregates 7980 individual prompts measured across 6262 monitoring runs over the past month.