Measured over the last 30 days
- Models Tracked
- 40
- Avg Tokens / Second
- 14.12
- Avg Time to First Token (ms)
- 12933.22
- Last Updated
- Sep 1, 2026
All venice models
40 tracked · fastest first| Model | Avg Toks/Sec | Min | Max | Avg TTF (ms) |
|---|---|---|---|---|
| Mistral Small 4 | 46.60 | 36.20 | 54.00 | 900.00 |
| Qwen3 Coder 480B A35B | 45.70 | 45.70 | 45.70 | 630.00 |
| DeepSeek V4 Flash 0731 | 44.80 | 44.80 | 44.80 | 1150.00 |
| Uncensored | 41.30 | 33.70 | 52.10 | 1060.00 |
| GLM 5.2 | 34.20 | 31.10 | 37.00 | 1340.00 |
| Uncensored | 33.50 | 16.20 | 59.50 | 1340.00 |
| Gemma 4 31B | 32.20 | 23.40 | 40.00 | 930.00 |
| Gemma 4 31B | 28.90 | 28.90 | 28.90 | 1030.00 |
| Mistral Small 3.2 24B | 24.80 | 14.20 | 34.20 | 1370.00 |
| Mistral Small 3.2 24B | 24.30 | 6.88 | 39.60 | 1190.00 |
| Gemma 4 26B A4B | 21.50 | 5.50 | 44.70 | 1130.00 |
| Qwen3 VL 235B A22B Instruct | 19.70 | 11.80 | 34.10 | 1340.00 |
| Qwen3.8 2.4T A95B | 17.10 | 13.00 | 21.10 | 3370.00 |
| MiniMax M3 | 16.50 | 14.70 | 18.80 | 3400.00 |
| Qwen3 VL 235B A22B Instruct | 16.10 | 3.73 | 30.90 | 1810.00 |
| Nemotron 3.5 Lightning | 14.30 | 9.80 | 20.00 | 4640.00 |
| Qwen3 235B A22B Instruct 2507 | 14.20 | 9.49 | 20.30 | 850.00 |
| Qwen3.5-35B-A3B | 10.90 | 8.79 | 12.40 | 5780.00 |
| Kimi K2.5 | 10.70 | 5.98 | 21.20 | 6710.00 |
| Qwen3.8 27B | 7.93 | 7.93 | 7.93 | 7580.00 |
| Qwen3 235B A22B Thinking 2507 | 7.66 | 5.59 | 9.25 | 7830.00 |
| Qwen3.6 35B A3B | 7.46 | 3.52 | 9.26 | 8790.00 |
| Kimi K2.7 Code | 7.30 | 2.73 | 11.70 | 8360.00 |
| DeepSeek V3.2 | 7.00 | 4.94 | 9.28 | 6300.00 |
| Qwen3.5-9B | 6.89 | 6.10 | 7.50 | 9020.00 |
| DeepSeek V4 Flash 0423 | 6.11 | 3.39 | 10.60 | 12100.00 |
| Qwen3.5-9B | 5.91 | 5.91 | 5.91 | 10300.00 |
| Qwen3.6 35B A3B | 5.76 | 3.18 | 8.14 | 12600.00 |
| MiniMax M2.5 | 5.13 | 4.60 | 5.98 | 11200.00 |
| GLM 4.7 Flash | 5.06 | 5.06 | 5.06 | 11800.00 |
| MiniMax M2.5 | 3.96 | 1.99 | 5.73 | 16900.00 |
| GLM 4.7 Flash | 3.83 | 2.45 | 4.75 | 17200.00 |
| Qwen3.5 397B A17B | 2.84 | 1.70 | 3.95 | 24200.00 |
| GLM 5.1 | 2.65 | 2.25 | 3.04 | 19900.00 |
| Kimi K2.6 | 2.34 | 0.99 | 3.96 | 35200.00 |
| GLM 5 | 2.20 | 0.97 | 3.79 | 36300.00 |
| DeepSeek V4 Pro 0423 | 1.59 | 1.59 | 1.59 | 21200.00 |
| GLM 4.7 | 1.23 | 0.60 | 2.14 | 65600.00 |
| GLM 4.6 | 0.99 | 0.79 | 1.36 | 66700.00 |
| GLM 4.6 | 0.75 | 0.61 | 0.91 | 84900.00 |
Frequently Asked Questions
Which venice model is fastest?
Based on recent tests, Mistral Small 4 shows the highest average throughput among tracked venice models.
How many recent measurements feed this dashboard?
This provider summary aggregates 5436 individual prompts measured across 4597 monitoring runs over the past month.