Niches · Cooking, recipes & meal planning
Estimate a meal's nutrition from a photo
No company's docs show a full fit yet. The listings below do the job with a limit, or not at all.
What counts as done
The dishes or ingredients recognised with estimated quantities and nutrients, and a confidence or a follow-up question where the photo is ambiguous, in one call.
Measured, for our own tests: the same photo gives totals within 10 per cent across three calls. No docs we read state these figures, so they don't count against a listing here.
How agents ask for it: what's in this plate and how many calories
, log this photo as lunch
Does the job, with a limit 4
It does the job, but something in the docs or a user's account will get in an agent's way. The limit is named in each row.
| Listing | Why |
|---|---|
| Edamam E Edamam | The Food Database plans include a Vision API (10,000 free calls on Basic, $0.015 a call on Core), but its response, confidence and retention were not documented on the pages read. source |
| fatsecret Platform API E fatsecret | image.recognition returns eaten_foods with suggested servings and food ids, but it is Premier Exclusive under the image-recognition scope, priced on request, and docs advise a confirm screen rather than auto-logging. source |
| LogMeal D AIGecko Technologies S.L. | segmentation/complete returns dishes per region with prob scores, and the quantity variant estimates grams on Recommend; each call logs an intake and spends a credit, so a retry double-logs, and photos are kept for model training by default. source |
| spoonacular D skycraft GmbH | food/images/analyze classifies the dish and guesses nutrition from 25 matching recipes with 95 per cent ranges, but returns one category with no ingredient quantities or follow-up question. source |
Doesn't do this job 2
Listings in the category that do something else, with the reason.
| Listing | Why |
|---|---|
| Eat This Much Partner API E Eat This Much, Inc. | No image endpoint in the OpenAPI document. source |
| Nutritionix API F Syndigo LLC | The Photo Upload API takes packaged-product photos (front, nutrition panel, ingredients) by UPC for the database; it does not recognise meals. source |
What stops agents finishing
- image recognition only on a quote-priced tier
- photos kept and used for training by default
- each retried call logs a second intake and spends a credit
Around it
Before: chat.user-uploads-photo. After: cooking.log-food.
Where this comes from
Weight 10 of 100 in Cooking, recipes & meal planning: 3 of 27 accounts, raised from 8 because three listings (LogMeal, fatsecret, Edamam) sell it and the carb-counting thread shows the repeatability problem
The same as data: /api/v1/use-cases/cooking.photo-to-nutrition.json, or the MCP tool get_use_case with id cooking.photo-to-nutrition. Other jobs in this category: /api/v1/niches?category=cooking.
For companies
Do agents find, use and choose your tools?
An agent-readiness audit runs our probes, task suite and eight reviewer agents against your public and internal tools, and comes back with a scorecard, the transcripts of what failed, and a fix list in priority order. From $2,500, re-run included. We never take payment to move a rank. We do help companies earn one.