This is documentation for Version 2 of the Food Database API. If you need to see what are the chnges from Version 1 please go to the Changelog

Food Database

This API provides you with tools to find nutrition and diet data for generic foods, packaged foods and restaurant meals. In addition it employs NLP (Natural Language Processing) which allows for extraction of food entities from unstructured text.

Covered Use Cases

• Search for a food by keyword, food name or UPC/Barcode
• Sourcing of nutrition facts for a given food, including: macro and micro nutrients, allergen labels, lifestyle and health labels
• Search for a food by given nutrient quantity for 28 nutrients
• Seach for foods within a given brand
• With the built in food-logging context it allows for NLP requests for chat bots and natural language calorie counters

Food Database Requests

Parse requests: https://api.edamam.com/api/food-database/v2/parser

Food Database Text Search

The parser access point handles text search for foods as well as filters for the foods like presence specific nutrient content or exclusion of allergens

• Search for a phrase or keyword using NLP to get food entities from it.
• Get basic nutrition facts and ingredients for each food
• Search for a food by given nutrient quantity for 28 nutrients
• Seach for foods within a given brand
• With the build in food-logging context it allows for requests which do not contain quantity and suggest expected quantities for them.

Parser Request

You will use a GET request to access the Parser. Either UPC or ingr should be present for a valid request

When searching by keyword

When using food logging context

If you use the food logging context feature it will modify the NLP response in the the following way
- You can submit items without quantity. Edamam will try to match them and assign quantity to them based on expected serving size
- The API will return as results only foods ready for direct consumption – no raw meats, raw dry goods, like raw rice for example
- Edamam can handle single items and two part compound items only – i.e. “chicken” or “rice AND chicken”. Make sure the URL is encoded properly

Pagination

To obtain the next page, the API user should follow the “next” link from the “_links” section in the result JSON, which looks like this:

"_links" : {
    "next" : {
      "title" : "Next page",
      "href" : "https://api.edamam.com/api/food-database/v2/parser?..."
    }
}

Search by nutrient range

When searching by keyword you can also specify nutrient ranges by adding parameters in the following form:

nutrients[NTR]=RANGE where

NTR is one of: CA, CHOCDF, CHOLE, FAMS, FAPU, FASAT, FAT, FATRN, FE, FIBTG, FOLDFE, K, MG, NA, NIA, P, PROCNT, RIBF, SUGAR, THIA, TOCPHA, VITA_RAE, VITB12, VITB6A, VITC, VITD, VITK1 or ZN;

and

RANGE is in one of MIN+, MIN-MAX or MAX, where MIN and MAX are non-negative integer numbers.

For example:
nutrients[CA]=50+ means minimum 50mg calcium, where ‘50+’ has to be properly encoded as ‘50%2B’
nutrients[FAT]=30 means maximum 30g fat and
nutrients[FE]=5-10 means iron between 5mg and 10mg inclusive

You can combine more then one nutrient ranges in a search requests

Example Parser request

As an example, let’s say we want to find matches in the food database for a red apple. We then need to URL-encode this string. In this case, this means to just replace the spaces with %20,so it becomes “red%20apple” Please note, that the quotation marks aren’t part of the string.
Here is an example using curl:

'https://api.edamam.com/api/food-database/v2/parser?ingr=red%20apple&app_id={your app_id}&app_key={your app_key}'

NOTE: Please make sure you use the credentials you created for this exact API as they are app and plan specific. The ${ } notation is to signify the type of input and should NOT to be included in the request itself.

With ‘food-logging’ turned on:

'https://api.edamam.com/api/food-database/v2/parser?nutrition-type=logging&ingr=red%20apple&app_id={your app_id}&app_key={your app_key}'

Parser Response

Parsed

NOTE: The ‘parsed’ section of the response contains the direct result from NLP handling of the request with it broken down into quantity/measure/food. This data then is used to produce the results from the ‘hints’ section.

Hints

Example response to text search

*Nutrition Data Requests

Nutrition Requests: https://api.edamam.com/api/food-database/v2/nutrients

In the response to your parser request you receive the a food ID for each database match. Using the food ID and the measure URI which parser provides you can make a request to the nutrients acces point. The nutrients access points returns nutrition with diet and health labels for a given quantity of the food.

Nutrients Request

The contents of the request must be a JSON object with the following format:

Ingredient

The Parser response for each food contains all specialized measures for this food. For example if an apple has a specific unit named ‘slice’ it will come with the response from the API.

In addition to the units provided by the API Edamam supports virtually any weight and volume measures for all foods. Edamam does not return them in the API as it makes the response redundant and cumbersome.

Here is a list of the standard supported measures which can be used in addition the the measures returned with the food:

A given measure may or may not contain contain a field “qualified”. This field contains possible measure qualifiers for the basic measure with each of them having it is own URI. For example for the item ‘apple’ one of the measures is ‘whole’ with qualifiers ‘large’, ‘small’ etc.

When submitted together with the measure URI the qualifier URI changes the weight of the basic measure.

The API returns nutritional analysis for the specified ingredient.

Each food comes with a list of specialized units which belong to it. For example if an apple has a specific unit named slice it will come with the response from the API.

Example Nutrients request

You will use a POST request to access ‘nutrients’.

Here is an example using curl:

This will send the food.json file for processing.

Here is the contents of the food.json file:

{
  "ingredients": [
    {
      "quantity": 1,
      "measureURI": "http://www.edamam.com/ontologies/edamam.owl#Measure_unit",
      "foodId": "food_bnbh4ycaqj9as0a9z7h9xb2wmgat"
    }
  ]
}

When a qualifier is available for a given measure it can be used in the following way:

{
  "ingredients": [
    {
      "quantity": 1,
      "measureURI": "http://www.edamam.com/ontologies/edamam.owl#Measure_unit",
      "foodId": "food_bnbh4ycaqj9as0a9z7h9xb2wmgat",
      "qualifiers": [ "http://www.edamam.com/ontologies/edamam.owl#Qualifier_large" ]
    }
  ]
}

Nutrients Response

Example Nutrients response

Nutirion

For ‘Nutritional Label Definitions’ consult the table at the bottom of this document

NutrientInfo

Ingredient

UPC or Barcode Search

Allows for search of an item based UPC/Barcode number.

This is a service allowing you to submit a UPC or barcode and find a match for it in the food database.

Here is an example using curl:
'https://api.edamam.com/api/food-database/v2/parser?upc={upc code}&app_id={your app_id}&app_key={your app_key}'

Autocomplete Request

Edamam provides a convinient autocomplete functionality which can be used when searching for ingredients.

Path: http://api.edamam.com/auto-complete

The endpoint returns suggestions for the text submitted to it. Here is an example

http://api.edamam.com/auto-complete?q=rou&limit=10&app_id=${YOUR_APP_ID}&app_key=${YOUR_APP_KEY}

Nutritional Label Definitions

Nutrition labels are shared by recipes and foods alike. They are assigned by Edamam based on the ingredients contained in the food label for CPG foods and by the basic ingredients of each recipe for recipes.

Types

Composite types are described in terms of their JSON representation.

Throughout descriptions, the following notation is used:

• integer, float, and string stand for the JavaScript primitive types integer, float, and string, respectively
• enum stands for a string field that only takes on values from a pre-defined range (the range is specified where essential)
• T[] stands for an array of objects of type T
• T[*] stands for an object (associative map) whose every field (element) is of type T.