Not sure where to start? Take a guided tour.
AI Site Grader
AI Query Scanner
llms.txt GeneratorAverage top-three overlap is 2.8
The average top-three result overlap across evaluated AI models is 2.8 in the model divergence benchmark.
The 'Same Question, Different AI, Different Answers' study evaluates the consistency of AI model outputs by measuring the average overlap among the top three results generated by different models.
The numbers behind it.
- Average top 3 overlap
- Average overlap among top-three results across models.
- 2.8
The evidence.
| Measure | Result |
|---|---|
| Average top 3 overlap Average overlap among top-three results across models. | 2.8 |
Population and method
Built from 797,644 valid comparisons across 44,088 reports and 8 models, covering 6,439,133 model responses in the observed window.
Where the evidence stops
Agreement is measured across recommendation outputs, not across hidden reasoning or retrieval context.
Average agreement can hide large differences between query classes and model pairs.
The study measures overlap, not which answer was objectively “right”.
What to do with this
Operators and strategists should interpret this 2.8 overlap score as an indicator that while consensus exists among category leaders, relying on a single model obscures critical differences in inclusion and ranking. Multi-model optimization is required to guarantee comprehensive visibility.
Cite this fact.
Source study: Model Divergence“Average top-three overlap across models is 2.8 in the model divergence benchmark.”
Same Question, Different AI, Different Answers - Trakkr Research · Average top-three overlap is 2.8
Same Question, Different AI, Different Answers · Trakkr Research · Updated 11 March 2026