Nourish just leveled up on Zap Cooking. π₯¦β‘
The foundation was already built. We just added estimated nutrition macros and made sure the infrastructure underneath it was sound.
Every analyzed recipe now includes estimated nutrition per serving:
π₯ Calories
πͺ Protein
πΎ Carbs
π₯ Fat
Nourish uses AI to recognize the ingredients and estimate the nutrition values. From there, the calculation is simple: add everything together and divide by the number of servings.
When a recipe is harder to estimate, we label it as a rough estimate instead of pretending the numbers are exact.
Guidance, not gospel.
We tested Nourish against a large set of recipes with published nutrition before releasing the update. Many of those recipes are now part of the Nourish catalog, so you can explore the results for yourself.
The code is open source, the method is transparent, and the data lives on Nostr.
Every analysis is published as a signed event to our Pantry relay and broadcast across multiple public relays. No account required. No nutrition data trapped inside a private database. Just portable, verifiable information built on an open protocol.
That data also makes discovery easier:
πͺ High protein
π₯ Under 600 calories
πΎ Low carb
π« No seed oils
π¬ No added sugar
Filters can be combined, so you can find meals that are high protein and free from seed oils in just a couple of taps.
Thanks to
@utxo the webmaster π§βπ» for the recommendation to add the nutrition layer. The foundation was there. We added the macros and made sure it was built the right way.
Try Nourish:
https://zap.cooking/nourish/exploreBuilders: analyses are published as kind 30078 events with NIP 32 labels and signed by the Zap Cooking service key.
Query:
wss://pantry.zap.cooking
No authentication required. Build something with it. That is the point.
Built open. Built on Nostr. Built the way it should be. πΏ
Food is Open Source, and so are we.