system-one-bench

~/system-one-bench / compare/baseten-mammoth-1-vs-nace-ai-drex-v1-5

Head to head

Mammoth 1 vs Drex v1.5

Mammoth 1 scores 82.9 and Drex v1.5 77.6 on the System One Index. Mammoth 1 is more accurate on 15 of 16 text datasets, Drex v1.5 on 0, and 1 are tied. Same cases, same request, same scorer.

Overall

Mammoth 1Drex v1.5
System One Index82.977.6
English82.577.6
Arabic82.877.0
Vision Indexโ€“โ€“
Calibration error4.9%10.8%
Median latencyโ€“198 ms
Cost per 1,000 calls$0.037$0.010

Text datasets, accuracy

Mammoth 1Drex v1.5
AG News English92.090.0
SST-2 English96.094.5
SST-5 English59.053.5
DAIR Emotion English59.053.5
Banking77 English94.587.0
BoolQ English92.089.0
XNLI English English91.388.7
Typed decisions English76.464.8
XNLI French French86.082.0
XNLI Arabic Arabic78.774.0
MASSIVE scenario Arabic75.058.5
MASSIVE intent Arabic91.077.5
OSACT4 offensive Arabic91.590.5
Tweet sentiment Arabic64.563.5
MIRACL rerank Arabic89.089.0
MIRACL pairwise Arabic90.086.0

Model pages: Mammoth 1 ยท Drex v1.5. Latency for API models is the median per call through OpenRouter, timed from a cloud server with 8 calls in flight.