OpenAI · GPT-5.6

GPT-5.6 Luna

The cost-efficient GPT-5.6 tier, optimized for broad routine deployment while retaining substantial family-level capability.

DOOM SCORE80.6out of 100

CURRENT ASSESSMENT · REVISION 1

Why this model scores 80.6

Very broad affordable access raises diffusion risk, while lower capability and autonomy than the larger GPT-5.6 tiers reduce the overall profile.

Capability84

Autonomy82

Deployment98

Misuse potential75

Control difficulty65

AUDIT TRAIL