How we measure answer quality
This page used to publish a quality score for Education and comparison figures for named competitors. Those numbers were not the result of a measurement we could show you, so we removed them. What replaced them is below.
Answers built on retrieved content are scored, for the customer who owns them
When the assistant answers from your material, that answer is scored for factuality and flagged if it states something the retrieved content does not support. Those results are visible to you for your own site, alongside the questions the assistant could not answer.
That is a real measurement, and it is the one that matters: how the assistant performs on your content, not an average across nine industries.
The first run is published
On 4 November 2026 we asked two of our demo assistants, the business demo and the medical demo, 120 fixed questions in English, Hebrew and Russian and scored every answer by written rules. The results, the weak spots included, are published with every raw answer, the questions and the scripts: answer quality, measured in public.
The Education vertical was not part of the 4 November 2026 run. The figures from that run describe those two demo sites, not this vertical, so we do not repeat them here.
We will not score a competitor's product on our own question set and publish the result as though it were neutral. Where we compare, we compare on facts you can verify against a vendor's own published terms: price, channels, languages, compliance posture.
Judge it on your own content
The free tier gives you 100 answers a month against your own material. That will tell you more than any scoreboard.