Model Scorecard.

Same question. Same settings. Different day. Different answer. A drift is when an AI model changes its mind, unprompted, unprovoked. This scorecard tracks how often each model does it.

All models run under identical conditions: fixed parameters, no randomness, yes-or-no only. Ranked by consistency.

Models

Daily at temp 0

426

Questions

82,788

Responses

1 run = 5 responses

5,243

Drifts

← swipe to see all models →

	Combined	Claude	GPT	Gemini	DeepSeek	Grok
Version	5 models	opus-4-6	gpt-5.4	2.5-pro	deepseek-v3	grok-4.2
Responses	82,788	16,639	16,639	16,639	16,639	16,639
Total Drifts	5,243	264	1,023	1,564	617	1,775
Questions Drifted	—	78	146	134	75	174
Drift Rate	6.3%	1.6%	6.1%	9.4%	3.7%	10.7%
Drift Per Day	163.8	8.2	32.0	48.9	19.3	55.5
Top Drifter	Q68	Q68	Q33	Q57	Q33	Q70
Stability Rank	—	#1	#3	#4	#2	#5

What each row means

Run: One question asked once across all five models
Responses: Number of times a model answered YES or NO. 1 run = 5 responses.
Total Drifts: How many times a model changed its answer (YES→NO or NO→YES)
Questions Drifted: How many unique questions a model has flipped on at least once
Drift Rate: Drifts as a percentage of total responses
Drift Per Day: Average number of answer changes per day
Top Drifter: The question this model has flipped on the most
Stability Rank: Models ranked #1 (most stable) to #5 (least stable) by total drifts