Trust & transparency

How we research and compare AI tools

How AI Product Reviews researches products, records evidence and distinguishes documentation from hands-on tests.

What our reviews establish

Our refreshed reviews are documentation-based research. They cite official product pages, help centres and pricing information, and separate vendor claims from editorial advice. They are not hands-on benchmarks unless a page explicitly describes a reproducible test and its results.

Dates and evidence

A research-checked date means the cited information was reviewed on that date. It is not the publication date or a claim that every feature was tested. Older imported overviews have no recorded check date and can contain outdated information. Check the vendor before buying.

No generated product scores

We do not display the numerical sorting values imported with the legacy directory as product ratings. Recently checked pages appear first in discovery lists, followed by alphabetical listings. List positions do not establish that one product performs better than another.

How to use a review

Compare the intended workflow, limitations, plan restrictions and alternatives. Where we could not verify a product or price, we say so. Practical evaluation suggestions are checks for your own trial, not results we measured.

Corrections and future testing

We prioritise historically useful pages for research updates. Hands-on reviews require an identified plan, test date, inputs, outputs and limitations before we claim measured performance. Send factual corrections through our contact page.