system-one-bench

~/system-one-bench / compare/baseten-joey-1-vs-inception-mercury-decide

Head to head

Joey 1 vs Mercury Decide

Joey 1 scores 73.3 and Mercury Decide 82.4 on the System One Index. Joey 1 is more accurate on 2 of 16 text datasets, Mercury Decide on 14. Same cases, same request, same scorer.

Overall

Joey 1Mercury Decide
System One Index73.382.4
English72.983.7
Arabic73.580.8
Vision Index–79.4
Calibration error2.1%13.1%
Median latency–276 ms
Cost per 1,000 calls$0.004Free

Text datasets, accuracy

Joey 1Mercury Decide
AG News English85.092.0
SST-2 English86.095.5
SST-5 English40.552.0
DAIR Emotion English52.591.5
Banking77 English93.086.0
BoolQ English80.090.5
XNLI English English84.091.3
Typed decisions English62.471.1
XNLI French French74.782.7
XNLI Arabic Arabic64.776.0
MASSIVE scenario Arabic71.572.0
MASSIVE intent Arabic84.585.0
OSACT4 offensive Arabic83.092.0
Tweet sentiment Arabic66.062.5
MIRACL rerank Arabic62.088.0
MIRACL pairwise Arabic83.090.0

Image datasets, accuracy

Joey 1Mercury Decide
SST-2, as an image–93.0
Arabic sentiment, as an image–39.0
Oxford-IIIT Pet–84.5
VQAv2 yes/no–87.0
CountBench–93.5

Model pages: Joey 1 · Mercury Decide. Latency for API models is the median per call through OpenRouter, timed from a cloud server with 8 calls in flight.