GPT-5.6 Luna (low reasoning effort)
OpenAI · proprietary API · closed / proprietary · llm-baseline
JevBench v1.3.0 score
65.9
Rank #14 of 48 ranked systems.
8.5 points behind Jev 1.13.0's 74.4.
Axes
- Intelligence
- 95.3
- Calibration
- 89.8
- Speed
- 77.5
- Cost
- 28.5
Cost
$0.242 per 1,000 decisions.
See every JevBench v1.3.0 system, method and scoring on the full leaderboard.