Measured over the last 30 days
- Models Tracked
- 10
- Avg Tokens / Second
- 20.52
- Avg Time to First Token (ms)
- 5574.00
- Last Updated
- Oct 16, 2026
| Model | Avg Toks/Sec | Min | Max | Avg TTF (ms) |
|---|---|---|---|---|
| GLM 5.2 | 60.10 | 60.10 | 60.10 | 670.00 |
| gpt-oss-120b | 38.90 | 30.70 | 47.10 | 1220.00 |
| Llama 3.3 70B Instruct | 36.30 | 9.68 | 61.00 | 1580.00 |
| MiniMax M3 | 18.90 | 5.75 | 36.60 | 3750.00 |
| Muse Glimmer 30B | 12.50 | 11.20 | 13.70 | 4610.00 |
| Qwen3.8 2.4T A95B | 12.50 | 12.50 | 12.50 | 4640.00 |
| Nemotron 3.5 Super VL | 8.65 | 8.65 | 8.65 | 7050.00 |
| DeepSeek V4.1 Flash | 6.52 | 6.52 | 6.52 | 9540.00 |
| DeepSeek V4 Flash 0731 | 5.97 | 5.97 | 5.97 | 9880.00 |
| Qwen3.5-9B | 4.84 | 4.84 | 4.84 | 12800.00 |
Frequently Asked Questions
Which together model is fastest?
Based on recent tests, GLM 5.2 shows the highest average throughput among tracked together models.
How many recent measurements feed this dashboard?
This provider summary aggregates 2416 individual prompts measured across 2116 monitoring runs over the past month.