Methodology

How we evaluate

Every score on this site comes from the same rubric, applied the same way, whether the tool pays a commission or not. Here's exactly how it works.

The four-part test

1. Documented research

We evaluate each tool against its actual public interface, documentation, and verified output, not its demo video or marketing copy. Every feature claim in a review is checked against what the product's own site and docs say it does, not what the landing page implies.

Hands-on testing is rolling out. We're building the capacity to run live test tasks inside each tool directly, and will label pages clearly once that's live for a given comparison. Until then, scores reflect documented research and verified pricing, not firsthand usage; we'd rather say that plainly than imply more than we've done.

2. Pricing math

We price every tier a real buyer would actually need, not just the entry price a vendor leads with. This includes credit systems, generation caps, seat minimums, mandatory onboarding fees, and overage charges: the line items that most comparison content leaves out because they're inconvenient, not because they don't matter.

3. Category fit

A tool built for a fifty-person marketing team is scored against that job, not against a solo creator's budget, and vice versa. Two tools solving different problems well both score highly; a tool solving the wrong problem for a stated use case scores lower even if it's a good product in general.

4. Ongoing recheck

Pricing and features are monitored on a weekly cadence. When a vendor changes pricing, discontinues a plan, or ships a feature that changes the comparison, the page gets updated, not left stale until someone complains. The "Last updated" date at the top of every page reflects the last time this actually happened, not just a cosmetic refresh.

The scoring scale

Every tool gets a score from 1 to 10, most useful in relative terms against the other tools on the same page rather than as an absolute universal ranking:

RangeWhat it means
9-10Best-in-class for its category, minimal real drawbacks at its price point
7-8.9Strong, clear recommendation for a specific use case or buyer profile
5-6.9Usable and honestly priced, but with real limitations worth knowing before buying
Below 5Included for comparison context; we wouldn't lead with it as a recommendation

Verdict badges

The verdict box at the top of every comparison page uses a small set of consistent badges:

Affiliate relationships and scoring

Some tools on this site are affiliate partners; some aren't yet (see each page's disclosure). Affiliate status is never a scoring input; the rubric above is applied identically either way, and every page explicitly discloses which links are affiliate links. Full policy on our privacy page and terms page.

Who does the work

An AI agent workforce runs the recurring research, pricing verification, and weekly recheck work described above, under a human-defined rubric and human review of every published verdict. More on that model on our about page.