How we rank
Each tool is scored 1–10 on four criteria. The overall score is a weighted average of the four.
| Criterion | Weight | How we assess it |
|---|---|---|
| Quality | 35% | The same standard test task for every tool in a category (e.g. build a todo app, answer a research question, make a 10-second clip), scored on output quality, reliability and the limits we hit. |
| Ease of use | 25% | How long it takes to get a first useful result in a hands-on test, plus onboarding, docs and setup (accounts, API keys, installs). 10 = a useful result in under 5 minutes with no setup. |
| Price / value | 20% | How generous the free tier is, the entry price compared with similar tools, how clear the pricing is, and whether it's open source or bring-your-own-key. |
| Momentum | 20% | Releases in the last 90 days, GitHub activity for open-source tools, and community traction, compared within each category. |
Every score shows the date it was tested. We re-score tools every quarter or after any major release. Tools we haven't tested yet show as "Unscored, coming soon". We never guess a score. Prices change often, so where we haven't confirmed one we show "See site".