AI Rotten Tomatoes: Honest Scores for AI Models
AI Rotten Tomatoes is the Rotten Tomatoes for AI. Instead of vendor marketing, we collect real user reviews from Reddit, Hacker News, V2EX, Zhihu, YouTube, Bilibili and more, and give every AI model a worth-it score.
How the score is calculated
Each model gets one number: the share of reviewers who would pay for it again. worth-it ÷ (worth-it + not-worth-it), shown as a percentage. The higher the score, the more people think it's worth the money. Data comes from real user reviews and community votes, updated weekly.
Why “Rotten Tomatoes”
Rotten Tomatoes uses a freshness score to tell you whether a movie is worth watching; we use the same logic to tell you whether an AI model is worth buying. We borrow the metaphor — but the data, the scores, and every review are our own.
Check the models you care about
FAQ
What is AI Rotten Tomatoes?
AI Rotten Tomatoes is the Rotten Tomatoes for AI — community votes and real reviews that tell you which AI models are worth buying and which are hype.
How is the worth-it score calculated?
worth-it score = worth it ÷ (worth it + not worth it) × 100. Data comes from real user reviews on Reddit, Hacker News, V2EX, Zhihu, YouTube, Bilibili, plus real on-site votes.
Which AI models are covered?
60+ mainstream models across Chat, Coding, Image, Video, Audio, Retrieval and Agent categories, updated weekly.
How is this different from vendor marketing?
Vendors sell capability ceilings; we tell you whether it's worth paying for — real usage experience, price and consensus, not specs and benchmarks.