Measured over the last 30 days
- Models Tracked
- 38
- Avg Tokens / Second
- 39.03
- Avg Time to First Token (ms)
- 6827.11
- Last Updated
- Aug 9, 2026
| Model | Avg Toks/Sec | Min | Max | Avg TTF (ms) |
|---|---|---|---|---|
| GPT-5.1-codex-mini | 86.00 | 17.10 | 105.00 | 990.00 |
| GPT-5 Nano | 83.90 | 8.99 | 159.00 | 2000.00 |
| GPT-5.1-codex | 69.00 | 20.30 | 83.40 | 1010.00 |
| GPT-5-chat-latest | 60.60 | 12.90 | 76.00 | 1140.00 |
| GPT-5.1-codex-max | 60.10 | 6.97 | 122.00 | 2910.00 |
| GPT-5-codex | 56.60 | 1.68 | 72.70 | 1570.00 |
| o3 Mini | 56.30 | 1.87 | 189.00 | 1850.00 |
| o1 | 51.90 | 5.22 | 142.00 | 2470.00 |
| GPT-5.1-chat-latest | 50.50 | 7.89 | 71.90 | 1700.00 |
| gpt-5.2-codex | 50.40 | 17.60 | 60.90 | 1290.00 |
| gpt-4.1-nano | 50.00 | 4.49 | 87.10 | 920.00 |
| gpt-4.1 | 49.30 | 4.80 | 78.70 | 780.00 |
| o4 Mini | 47.90 | 9.02 | 170.00 | 1680.00 |
| GPT-5 Mini | 47.80 | 4.46 | 112.00 | 2590.00 |
| GPT-5.4-nano | 46.70 | 13.90 | 90.40 | 1380.00 |
| GPT-5.4-mini | 42.90 | 5.28 | 76.20 | 1330.00 |
| gpt-3.5-turbo | 41.90 | 3.68 | 78.10 | 1200.00 |
| o3 | 41.40 | 9.19 | 104.00 | 2320.00 |
| GPT-5.2-chat-latest | 38.60 | 1.70 | 53.00 | 2190.00 |
| GPT-5 | 37.80 | 3.12 | 73.50 | 4120.00 |
| gpt-4o | 37.60 | 5.63 | 63.30 | 990.00 |
| gpt-4o-mini | 37.50 | 6.38 | 56.60 | 980.00 |
| gpt-5.3-chat-latest | 36.10 | 4.83 | 50.40 | 2430.00 |
| gpt-4.1-mini | 35.80 | 1.02 | 57.70 | 1160.00 |
| GPT-5.1 | 31.60 | 3.06 | 58.50 | 1720.00 |
| GPT-5.6-luna | 27.10 | 2.78 | 46.70 | 2180.00 |
| GPT-5.4 | 25.00 | 4.28 | 47.20 | 1850.00 |
| GPT-5.6-terra | 24.00 | 2.20 | 37.00 | 2050.00 |
| GPT-5.2 | 22.80 | 8.61 | 37.40 | 1900.00 |
| gpt-4 | 22.70 | 3.20 | 36.80 | 1390.00 |
| GPT-5.5 | 22.00 | 4.00 | 31.90 | 1990.00 |
| GPT-5.3-codex | 18.20 | 1.92 | 36.70 | 2220.00 |
| o3-pro | 16.90 | 3.57 | 36.40 | 26500.00 |
| GPT-5.6-sol | 16.50 | 3.96 | 27.70 | 2730.00 |
| GPT-5.2-pro | 14.50 | 5.25 | 22.30 | 30000.00 |
| GPT-5.5-pro | 11.60 | 0.67 | 35.40 | 25900.00 |
| GPT-5-pro | 9.73 | 7.65 | 11.40 | 36500.00 |
| GPT-5.4-pro | 3.91 | 1.98 | 9.84 | 81500.00 |
Frequently Asked Questions
Which openai model is fastest?
Based on recent tests, GPT-5.1-codex-mini shows the highest average throughput among tracked openai models.
How many recent measurements feed this dashboard?
This provider summary aggregates 20742 individual prompts measured across 18576 monitoring runs over the past month.