Food Intelligence API for B2B Applications
ActualPlate is a food-image analysis API being built for software teams. Send a photo, get back structured food and nutrition data your product can use — no in-house computer-vision or food-database work required.
API in active development. No production traffic is served yet.
POST /v1/analyze Content-Type: application/json { "image": "base64_or_url..." } 200 OK { "foods": [ { "name": "Grilled chicken breast", "quantity": { "value": 150, "unit": "g" }, "nutrition": { "calories": 220 } } ] }
Understanding food from a photo is harder than it looks
Any product that deals with food — nutrition, fitness, meal planning, wellness — eventually needs to turn a photo into usable data: what food is this, how much of it, and what does it contain. Building that in-house means computer vision, food databases, quantity estimation, and nutrition mapping, on top of the product you're actually trying to ship.
- Computer vision. Detecting and segmenting food in an image reliably.
- Quantity estimation. Turning a 2D photo into a plausible portion size.
- Nutrition mapping. Connecting identified foods to structured nutrition data.
- Ongoing maintenance. Keeping models and data current as food and use cases change.
One request, structured food data back
ActualPlate is being built around a straightforward pipeline. The steps below describe the planned workflow — the underlying architecture is still being finalized.
Food image
Your application sends a photo of food to the API.
Food detection
Individual food items are located and segmented in the image.
Quantity estimate
Approximate weight or portion size is estimated per item.
Nutrition mapping
Detected foods are mapped to structured nutrition data.
JSON response
A structured response is returned to your application.
A developer-first response shape
Below is a realistic example of the request and response format ActualPlate is being designed around. Values shown are illustrative — they are not measured production accuracy.
{
"image": "..."
}
{
"foods": [
{
"name": "Example food",
"quantity": {
"value": 150,
"unit": "g"
},
"nutrition": {
"calories": 220
}
}
]
}
Teams building food-adjacent products
ActualPlate is designed to sit behind other people's products as an API, not to be a consumer app itself.
Nutrition apps
Let users log food by photo instead of manual search.
Fitness platforms
Add food logging alongside activity and training data.
Meal-planning tools
Turn a photo into structured ingredients and quantities.
Wellness platforms
Give users a lightweight way to capture what they eat.
Food-tech products
Add image-based food understanding to an existing product.
Other applications
Any product that needs to interpret food images programmatically.
Built as an API, not a product wrapper
ActualPlate is being designed for integration from day one: a REST API returning structured JSON, meant to be called from your backend or app.
- REST API In development
- JSON responses In development
- Documentation Coming soon
- SDKs Planned
$ curl -X POST https://api.actualplate.com/v1/analyze \ -H "Content-Type: application/json" \ -d '{"image": "..."}' # Illustrative call — endpoint not yet live. # API-key authentication is planned; final # auth method to be confirmed at launch.
What we're building toward
These are the goals shaping the product — not claims about performance that hasn't been measured or published yet.
Built for B2B integration, not a standalone consumer app.
Structured, predictable API output your code can rely on.
Designed specifically for food-image workflows.
Nutrition-oriented output mapped from detected foods.
Developer-focused integration and documentation.
ActualPlate API is currently in development.
We're inviting a small number of companies building food, nutrition, fitness, or wellness products to get involved early.
- Discuss your use case with the team
- Evaluate whether ActualPlate fits your workflow
- Give feedback that shapes the API
- Get launch updates as the API develops