logo
Phytelix
Prevent • Detect • Cure

Phytelix API

Analyze crop images with The Phytelix models and fetch treatment guidance. Built for simple integration with clear JSON responses.

CORS is enforced. Your origin must be in ALLOWED_ORIGINS on the api.

Quick start

// Analyze an image
fetch(`${baseUrl}/analyze`, {
  method: 'POST',
  headers: { 'Content-Type': 'application/json' },
  body: JSON.stringify({
    image_base64: 'data:image/jpeg;base64,...',
    model_type: 'autodetect' // or 'corn', 'tomato', etc.
  })
}).then(r => r.json()).then(console.log);

// Get treatment
fetch(`${baseUrl}/treatment`, {
  method: 'POST',
  headers: { 'Content-Type': 'application/json' },
  body: JSON.stringify({ disease: 'Late Blight', crop: 'tomato' })
}).then(r => r.json()).then(console.log);

Endpoints

POST /analyze

Image → Predictions

Runs a real model on your base64 image and returns the top prediction plus all class probabilities.

Request body

{
  "image_base64": "data:image/jpeg;base64,<base64>",
  "model_type": "autodetect" // optional: corn | tomato | cassava | cacao | guava | apple | banana | orange
}

Tip: convert a File to base64 with FileReader in the browser.

Response body

{
  "disease_detected": "Late Blight",
  "confidence": 92,
  "predictions": [
    { "className": "Late Blight", "probability": 92 },
    { "className": "Early Blight", "probability": 7 }
  ],
  "model_used": "https://.../models/tomato/",
  "analysis_timestamp": "2025-09-28T16:34:22.120Z"
}

POST /treatment

Advice

Returns human‑readable treatment guidance for a given disease and crop.

Request body

{
  "disease": "Late Blight",
  "crop": "tomato"
}

Response body

{
  "response": "Immediate Actions for Late Blight...\n• ...",
  "disease": "Late Blight",
  "crop": "tomato"
}

Usage

Quota

GET /usage?user_id=YOUR_ID

{ "remaining_analyses": 99 }

POST /usage/decrement

{ "user_id": "YOUR_ID" } → { "remaining_analyses": 98 }

Backed by D1 (if configured) with a user_usage table; otherwise defaults to 100 remaining.

Model & History

  • GET /api/model-urls → JSON map of model types to AI base URLs.
  • GET /api/history → Recent diagnoses (if DB is configured).
  • POST /api/save → Save a diagnosis row. Body: { crop, disease, user? }.

CORS & Auth

Requests must originate from an allowed website:

const ALLOWED_ORIGINS = [
  // e.g. 'https://your-site.example', 'http://localhost:5173'
];

function getCorsHeaders(request) {
  const origin = request.headers.get('Origin');
  if (origin && ALLOWED_ORIGINS.includes(origin)) {
    return {
      'Access-Control-Allow-Origin': origin,
      'Access-Control-Allow-Methods': 'GET,POST,OPTIONS',
      'Access-Control-Allow-Headers': 'Content-Type,Authorization'
    };
  }
  return null;
}

Request access to model API by inputting your website url, and why you want access in an essay form. Send here

Playground

Test /analyze


          

        

        
        

          
Test /treatment

          

            

              
              
            

            

              
              
            

            
            

          

        

      

    


    
    

      

        Need help? Email support.