Quickstart

This guide gets you from an API key to a full TRA tax estimate in a single request.

1. Make a request

Send a supported listing URL to POST /v1/estimate:

curl -X POST https://import.kopagari.com/v1/estimate \
  -H "Authorization: Bearer tra_live_xxx" \
  -H "Content-Type: application/json" \
  -d '{ "url": "https://www.beforward.jp/toyota/vitz/sc123456/" }'

const res = await fetch("https://import.kopagari.com/v1/estimate", {
  method: "POST",
  headers: {
    "Authorization": `Bearer ${process.env.TRA_API_KEY}`,
    "Content-Type": "application/json",
  },
  body: JSON.stringify({ url: "https://www.beforward.jp/toyota/vitz/sc123456/" }),
});
const data = await res.json();
console.log(data.tax.totalTaxesInTZS);

import os, requests

res = requests.post(
    "https://import.kopagari.com/v1/estimate",
    headers={"Authorization": f"Bearer {os.environ['TRA_API_KEY']}"},
    json={"url": "https://www.beforward.jp/toyota/vitz/sc123456/"},
)
print(res.json()["tax"]["totalTaxesInTZS"])

2. Read the response

{
  "input": { "url": "https://www.beforward.jp/toyota/vitz/sc123456/" },
  "car": {
    "make": "TOYOTA", "model": "VITZ", "modelCode": "SJ5",
    "year": 2017, "engine_cc": 1490, "fuel": "PETROL",
    "fobPriceUSD": 8500, "cfPriceUSD": 11234
  },
  "matched": {
    "make": "TOYOTA", "makeId": 1500298,
    "modelBody": "VITZ - SJ5 - SEDAN",
    "matchType": "exact_code", "confidence": "high",
    "adjustments": [],
    "yom": 2017, "country": "JAPAN",
    "fuelType": "PETROL", "capacityBand": "1001 - 1500 CC"
  },
  "tax": {
    "cifInUSD": "12,345.00",
    "importDutyInUSD": "3,086.25",
    "exiseDutyInUSD": "771.56",
    "vatInUSD": "2,880.45",
    "totalImportTaxesInUSD": "6,738.26",
    "totalTaxesInTZS": "18,056,140.50",
    "vehicleRegistrationFeeInTZS": "500,000.00"
  },
  "cached": false
}

3. Trust the result with `confidence`

Each estimate reports how the vehicle was matched against the TRA database:

confidence: "high" — chassis code matched a TRA entry exactly, no adjustments needed.
confidence: "medium" — matched with minor field adjustments (e.g. fuel naming).
confidence: "low" — matched by AI, or the model/year had to be substituted.

The matched.adjustments array explains every value that differed from the listing — for example, when TRA has no entry for the exact year and the nearest year was used.

Estimates are cached for 14 days per URL. A repeated request returns instantly with "cached": true.

Getting started

1. Make a request

2. Read the response

3. Trust the result with `confidence`

Next steps

API Reference

Authentication

​1. Make a request

​2. Read the response

​3. Trust the result with confidence

​Next steps

API Reference

Authentication

1. Make a request

2. Read the response

3. Trust the result with `confidence`

Next steps