How we rank

What the letters mean

• S — Start here. The current default for that job.
• A — Strong alternative. Excellent, and sometimes the better fit for a specific stack.
• B — Narrower fit. Solid, less proven, or aimed at a smaller slice of the job.
• C — Behind. Usable, but early, stale, or a step back from the tools above it.

What the current lists are

Every placement in this catalog has tier_status: provisional. The tiers come from public evidence checked on the verification date: documented capabilities, vendor backing, pricing pages, and GitHub activity where a repo is listed. They are marked for hands-on testing. A blank field on a tool page means we did not verify it. We do not fill gaps with guesses.

The same tool can hold different letters on different lists. ElevenLabs can lead character voices and sit lower for game music. Read the rationale on that list, not a single global score.

The test we are building toward

This formula is the target. It is not how the seed tiers were calculated. Hands-on results will be published on the tool page before a placement leaves provisional status.

1. Hands-on test

   40%

   A fixed task for the category: did it work, how long, what it cost.

2. Momentum

   20%

   Public activity such as release pace and attention, cited when we use it.

3. Community sentiment

   20%

   What builders say in public, with the source linked.

4. Value and freshness

   20%

   Price, free tier, and whether the project is still alive.

What never changes a rank

• Paying for a listing, a newsletter slot, or a faster review.
• Affiliate links. If we add them, they stay disclosed and off the score.
• Sponsored drops. Those sit in the feed with a label, outside the tier list.

Submitting a tool is free. There is no paid skip-the-queue. Send a tool and it waits with everything else.