Measured over the last 30 days
- Models Tracked
- 5
- Avg Tokens / Second
- 28.74
- Avg Time to First Token (ms)
- 3150.00
- Last Updated
- Oct 16, 2026
All mara models
5 live · one row per model · fastest first| Model | Avg Toks/Sec | Min | Max | Avg TTF (ms) |
|---|---|---|---|---|
| gpt-oss-120b | 57.60 | 49.10 | 63.50 | 1140.00 |
| MiniMax M3 | 31.10 | 9.40 | 42.80 | 2700.00 |
| MiniMax M2.7 | 19.40 | 12.60 | 30.30 | 4330.00 |
| DeepSeek V3.2 | 18.30 | 8.31 | 34.80 | 4420.00 |
| DeepSeek V3.1 | 17.30 | 17.30 | 17.30 | 3160.00 |
Frequently Asked Questions
Which mara model is fastest?
Based on recent tests, gpt-oss-120b shows the highest average throughput among tracked mara models.
How many recent measurements feed this dashboard?
This provider summary aggregates 737 individual prompts measured across 718 monitoring runs over the past month.