About

We study the world's recipes. Then we build the best version.

The internet has millions of recipes, but the best version of a dish is still a guessing game. Ratings do not compare fairly across sites. Lists chase clicks. Nobody does the full research for you.

GüdDish was built to do that work: search the field, measure the evidence, learn what makes the strongest recipes great, and build one definitive GüdDish you can cook.

Not another recipe site

We do not host other people's recipes as if they were ours. On the Rank tab we publish the order and link to each publisher for the method. On the Cook tab we publish The GüdDish: our synthesis of what the field taught us. Publishers did the original cooking. We did the research.

Three beliefs

1. The best version is research, not luck

Someone should study the world's methods for a dish and write the one worth cooking. That is the job.

2. Evidence beats popularity

Stars alone are not enough. GüdScore finds the recipes with the strongest evidence so we know what deserves deep study. Then GüdDish builds from that learning, not from a single #1 reprint.

3. Creators still own their recipes

We do not replace publishers. We study them, rank them fairly, and link out. The GüdDish is our original synthesis. Their full methods stay on their sites.

GüdScore and GüdDish

GüdScore is the evidence engine. It analyzes thousands of recipes and millions of ratings so we can see which methods stand apart under honest measurement.

GüdDish is the product of that work: search, measure, learn, build. How the full pipeline works, what we exclude, and how GüdScore is computed live on How GüdDish works.

What we leave out of the score

If we cannot check the evidence behind a rating, we do not quietly include it in GüdScore. We would rather show a gap than invent certainty. Unrated methods and video can still inform technique when they teach something real. See Coverage for what we scanned and what we could not verify.

Where this started

GüdDish began with a familiar question: among hundreds of versions of a dish, which one is actually worth cooking? That question grew into a bigger one. What if we studied all of them and built the best version ourselves? That is still the product. One dish at a time.

Browse dishes →