Head to head
Drex Max scores 84.1 and Solar Decide 68.8 on the System One Index. Drex Max is more accurate on 15 of 16 text datasets, Solar Decide on 1. Same cases, same request, same scorer.
| Drex Max | Solar Decide | |
|---|---|---|
| System One Index | 84.1 | 68.8 |
| English | 82.8 | 64.7 |
| Arabic | 84.5 | 73.7 |
| Vision Index | – | – |
| Calibration error | 6.2% | 20.3% |
| Median latency | 277 ms | 715 ms |
| Cost per 1,000 calls | $0.021 | $0.042 |
| Drex Max | Solar Decide | |
|---|---|---|
| AG News English | 90.5 | 86.0 |
| SST-2 English | 96.0 | 91.5 |
| SST-5 English | 59.0 | 53.5 |
| DAIR Emotion English | 65.0 | 54.0 |
| Banking77 English | 90.5 | 1.0 |
| BoolQ English | 93.0 | 87.0 |
| XNLI English English | 95.3 | 74.7 |
| Typed decisions English | 73.0 | 69.9 |
| XNLI French French | 91.3 | 67.3 |
| XNLI Arabic Arabic | 89.3 | 58.0 |
| MASSIVE scenario Arabic | 73.0 | 55.5 |
| MASSIVE intent Arabic | 88.0 | 75.0 |
| OSACT4 offensive Arabic | 93.5 | 86.0 |
| Tweet sentiment Arabic | 69.0 | 71.5 |
| MIRACL rerank Arabic | 89.5 | 84.5 |
| MIRACL pairwise Arabic | 89.5 | 85.5 |
Model pages: Drex Max · Solar Decide. Latency for API models is the median per call through OpenRouter, timed from a cloud server with 8 calls in flight.