Measured over the last 30 days
- Models Tracked
- 11
- Avg Tokens / Second
- 21.29
- Avg Time to First Token (ms)
- 10427.33
- Last Updated
- Sep 1, 2026
All crusoe models
11 tracked · fastest first| Model | Avg Toks/Sec | Min | Max | Avg TTF (ms) |
|---|---|---|---|---|
| Nemotron 3 Nano 30B A3B | 89.10 | 86.20 | 92.10 | 550.00 |
| Nemotron 3 Nano 30B A3B | 72.10 | 49.20 | 88.30 | 750.00 |
| Llama 3.3 70B Instruct | 45.40 | 24.40 | 58.40 | 840.00 |
| Gemma 4 31B | 35.90 | 25.30 | 54.40 | 560.00 |
| GLM 5.2 | 34.10 | 19.90 | 45.60 | 1410.00 |
| DeepSeek V3 0324 | 30.20 | 8.63 | 46.00 | 1280.00 |
| GLM 5.2 | 29.60 | 8.17 | 43.40 | 2650.00 |
| DeepSeek V3 0324 | 29.30 | 14.00 | 37.10 | 1110.00 |
| Llama 3.3 70B Instruct | 21.40 | 19.90 | 22.80 | 1510.00 |
| Kimi K2.6 | 8.25 | 1.15 | 18.10 | 22600.00 |
| GLM 5.1 | 5.02 | 4.75 | 5.38 | 12400.00 |
Frequently Asked Questions
Which crusoe model is fastest?
Based on recent tests, Nemotron 3 Nano 30B A3B shows the highest average throughput among tracked crusoe models.
How many recent measurements feed this dashboard?
This provider summary aggregates 2981 individual prompts measured across 2820 monitoring runs over the past month.