System One Bench

Head to head

GPT-6 Luna Decisions vs Solar Decide Flash

GPT-6 Luna Decisions scores 76.0 and Solar Decide Flash 63.6 on the System One Index. GPT-6 Luna Decisions is more accurate on 15 of 16 text datasets, Solar Decide Flash on 1. Same cases, same request, same scorer.

Overall

GPT-6 Luna DecisionsSolar Decide Flash
System One Index76.063.6
English75.261.3
Arabic78.866.5
Vision Index81.6–
Calibration error9.8%14.4%
Median latency215 ms586 ms
Cost per 1,000 calls$0.040$0.042

Text datasets, accuracy

GPT-6 Luna DecisionsSolar Decide Flash
AG News English87.586.0
SST-2 English93.590.5
SST-5 English61.544.5
DAIR Emotion English58.554.5
Banking77 English77.51.0
BoolQ English85.586.0
XNLI English English66.765.3
Typed decisions English71.062.5
XNLI French French62.061.3
XNLI Arabic Arabic54.052.0
MASSIVE scenario Arabic66.046.0
MASSIVE intent Arabic84.563.0
OSACT4 offensive Arabic93.584.5
Tweet sentiment Arabic80.065.5
MIRACL rerank Arabic86.071.5
MIRACL pairwise Arabic87.583.0

Image datasets, accuracy

GPT-6 Luna DecisionsSolar Decide Flash
SST-2, as an image88.5–
Arabic sentiment, as an image67.0–
Oxford-IIIT Pet87.0–
VQAv2 yes/no80.5–
CountBench85.0–

Model pages: GPT-6 Luna Decisions · Solar Decide Flash. Latency for API models is the median per call through OpenRouter, timed from one machine.