Not sure where to start? Take a guided tour.
AI Site Grader
AI Query Scanner
llms.txt GeneratorOnly 4.0% of prompts produce perfect consensus
Only 4.0 percent of identical prompts produced perfect agreement across all eight analyzed artificial intelligence models.
An analysis of eight major artificial intelligence models including OpenAI, Anthropic, Gemini, Grok, Deepseek, Meta, Perplexity, and Google AI Overviews reveals significant variance in output generation. When presented with identical prompts, the models generated unanimous responses in only a marginal fraction of cases.
The numbers behind it.
- Perfect agreement
- Only a small share of prompts produce unanimous outcomes.
- 4.0%
- Models analyzed
- OpenAI, Anthropic, Gemini, Grok, Deepseek, Meta, Perplexity, and Google AI Overviews.
- 8
The evidence.
| Measure | Result |
|---|---|
| Perfect agreement Only a small share of prompts produce unanimous outcomes. | 4.0% |
| Models analyzed OpenAI, Anthropic, Gemini, Grok, Deepseek, Meta, Perplexity, and Google AI Overviews. | 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
Strategists and operators must diversify their artificial intelligence testing environments. Relying on a single model to validate search visibility or output accuracy is insufficient, as a single model win does not indicate broad market leadership or consistent cross platform performance.
Cite this fact.
Source study: Model Divergence“Only 4.0% of prompts produced perfect agreement across all 8 models.”
Same Question, Different AI, Different Answers - Trakkr Research · Only 4.0% of prompts produce perfect consensus
Same Question, Different AI, Different Answers · Trakkr Research · Updated 11 March 2026