Measured over the last 30 days
- Models Tracked
- 15
- Avg Tokens / Second
- 23.34
- Avg Time to First Token (ms)
- 6944.00
- Last Updated
- Oct 14, 2026
All cloudflare models
15 live · one row per model · fastest first| Model | Avg Toks/Sec | Min | Max | Avg TTF (ms) |
|---|---|---|---|---|
| Llama 3.2 3B Instruct | 72.60 | 65.20 | 80.10 | 580.00 |
| Gemma 4 26B A4B | 42.70 | 15.20 | 74.00 | 800.00 |
| Llama 3.2 1B Instruct | 41.00 | 22.30 | 59.70 | 510.00 |
| Granite 4.0 Micro | 28.90 | 28.70 | 29.20 | 560.00 |
| Llama 3.3 70B Instruct | 28.90 | 6.56 | 45.80 | 1670.00 |
| Mistral Small 3.1 24B | 26.90 | 22.80 | 31.10 | 870.00 |
| Qwen2.5 Coder 32B Instruct | 25.60 | 25.30 | 25.90 | 590.00 |
| GLM 5.2 | 15.80 | 3.04 | 29.10 | 4540.00 |
| Kimi K2.7 Code | 15.50 | 4.67 | 29.60 | 4310.00 |
| Llama 3.1 8B Instruct | 14.80 | 14.80 | 14.80 | 1220.00 |
| DeepSeek V4 Flash 0731 | 14.70 | 6.21 | 23.60 | 4200.00 |
| DeepSeek V4 Flash 0423 | 13.90 | 7.75 | 20.80 | 4410.00 |
| Kimi K2.6 | 4.51 | 2.25 | 9.58 | 14900.00 |
| DeepSeek V4 Pro 0423 | 2.66 | 1.37 | 4.54 | 27700.00 |
| GLM 4.7 Flash | 1.64 | 1.64 | 1.64 | 37300.00 |
Frequently Asked Questions
Which cloudflare model is fastest?
Based on recent tests, Llama 3.2 3B Instruct shows the highest average throughput among tracked cloudflare models.
How many recent measurements feed this dashboard?
This provider summary aggregates 1810 individual prompts measured across 1645 monitoring runs over the past month.