How we score and rank

Can you see a price first 30%
Whether the platform publishes what it costs before you register. Almost none of this category does, and that is the single most useful thing to know about it.
What the free tier actually does 25%
Whether you can browse, match and message without paying, or whether the free account is a shop window that stops at the first message.
How it bills you 20%
Subscription, credits or freemium. Credits mean the operator is paid per message rather than per outcome, which changes the incentive and makes your spend hard to predict.
Safety and verification 15%
What identity or profile verification the platform states it performs, and what it says about reporting and blocking.
How hard it is to leave 10%
Whether renewal terms and cancellation are stated plainly and in advance rather than discovered afterwards.

What this is

A terms-and-pricing comparison, not a field test. We have not dated our way through these services and we do not pretend to - claiming otherwise would be the easiest thing to write and the hardest thing for you to check.

What the scores are built from

What each platform publishes on its own site: pricing pages, free-tier descriptions, stated verification practices and published subscription terms. Every entry is dated. Where a platform publishes nothing on a point, the entry says "not published" and that counts against it - refusing to state a price before registration is information about how the business works.

What we deliberately do not do

We do not rank on match quality or success rates. Nobody can verify those, the platforms' own figures are marketing, and a review site that scores them is inventing a measurement.

How the score is calculated

Every platform gets a mark out of 10 on each of the five criteria above. The overall score is those five marks multiplied by their weights and added up - nothing else goes into it. We do not store the total anywhere; it is calculated from the individual marks each time the page is built, and the ranking order is simply the result sorted.

That matters more than it sounds. A headline score typed in by hand can be nudged, and nobody reading the page could tell. Here there is nothing to nudge: change a position and you have to change one of the five marks, which is a claim about a specific, checkable fact.