How we rate
Same five criteria on every card, weighted by category, checked against the current version, and dated.
1. Five criteria, out of 10
Every tool card scores five things: capability (how good is it at the job it claims), ease of use (time to first useful result), value (what the free tier gives you and what the paid tiers cost against rivals), trust (privacy terms, data-training defaults, business stability), and ecosystem (integrations, platforms, exportability). The weights change by category and the card says so: ecosystem counts more for a chatbot than for a voice generator.
2. Current versions, from primary sources
Ratings reflect each tool's current documented capabilities, pricing and terms, drawn from the vendor's own plan pages and release notes, plus our own sessions and major independent reviews where the card says so. We do not run a formal lab benchmark and we do not pretend to.
3. Dated, and re-checked
Every card carries the month it was last checked. Flagship tools are re-checked whenever they ship a major update, and every card is re-checked at least quarterly. When a rating changes, the card keeps the old score visible with its date and says what changed.
4. Suits, not silos
Cards are grouped into four suits so you can compare like with like: Spades for chat and assistants, Hearts for creative tools (image, voice, video, music), Diamonds for productivity and business, and Clubs for developer tools. A 9 in Hearts is not a 9 in Clubs; ratings compare within a suit.
5. Money never scores
Some vendor links are affiliate links and are marked as such. They do not affect ratings, ranks, or verdicts, and a tool with no affiliate program is treated exactly the same as one with a generous one. Details in our disclosure.