The Litmus Lab.
Testing AI Under Constraint.
← swipe to navigate →
#148
AI Models
Is AI anything more than a sophisticated auto-complete?
Added 2026-02-11 · 58 runs · 22 drifts
Claude
claude-opus-4-6
YES
80% yes
GPT
gpt-5.4
YES
100% yes
Gemini
gemini-2.5-pro
YES
100% yes
DeepSeek
deepseek-chat (V3.2)
YES
94% yes
Grok
grok-4.2-beta
YES
74% yes
Full Record