Head to head
d1 scores 82.7 and Microsoft-Decision-1 82.0 on the System One Index. d1 is more accurate on 10 of 16 text datasets, Microsoft-Decision-1 on 4, and 2 are tied. Same cases, same request, same scorer.
| d1 | Microsoft-Decision-1 | |
|---|---|---|
| System One Index | 82.7 | 82.0 |
| English | 81.3 | 81.0 |
| Arabic | 83.9 | 83.2 |
| Vision Index | – | – |
| Calibration error | 2.9% | 2.8% |
| Median latency | 188 ms | 240 ms |
| Cost per 1,000 calls | $0.015 | $0.010 |
| d1 | Microsoft-Decision-1 | |
|---|---|---|
| AG News English | 90.0 | 89.0 |
| SST-2 English | 95.5 | 95.5 |
| SST-5 English | 56.5 | 62.5 |
| DAIR Emotion English | 63.5 | 57.0 |
| Banking77 English | 89.0 | 87.0 |
| BoolQ English | 90.0 | 88.5 |
| XNLI English English | 91.3 | 89.3 |
| Typed decisions English | 74.5 | 79.0 |
| XNLI French French | 84.7 | 82.0 |
| XNLI Arabic Arabic | 80.0 | 72.0 |
| MASSIVE scenario Arabic | 73.0 | 83.5 |
| MASSIVE intent Arabic | 90.0 | 89.0 |
| OSACT4 offensive Arabic | 95.5 | 88.5 |
| Tweet sentiment Arabic | 67.5 | 68.5 |
| MIRACL rerank Arabic | 90.5 | 90.0 |
| MIRACL pairwise Arabic | 91.0 | 91.0 |
Model pages: d1 · Microsoft-Decision-1. Latency for API models is the median per call through OpenRouter, timed from one machine.