AI Recommendation Index
Which vendors do AI assistants actually name — and how stable is the answer?
When a buyer asks a general-purpose AI assistant which product to use, which names come back, how often, and do they hold still when the same question is asked again?
A recurring measurement of vendor recommendation in general-purpose AI assistants. Each edition freezes a prompt set, samples each prompt several times across several models, and records every vendor named and which one was named first. We publish the full result table, the prompt list and the raw scoring as CSV and JSON so anyone can reproduce or contest the numbers. The index measures what assistants say, not what is best.
- Q3 2026 · v1.0 · 18 Sep 2026
AI Recommendation Index — Infrastructure, Q3 2026
We put 12 buying questions about vector databases, agent frameworks and inference providers to three models, three times each — 108 sampled answers — and recorded every vendor named. Two findings stand out: recommendation is heavily concentrated at the top of each category, and the first name an assistant gives is unstable. In 36% of model-prompt cells, the same model asked the identical question three times led with a different vendor each time.