Get a Demo

API directory · Google Maps Platform

Geocoding API: address-to-coordinate conversion for India products

The Geocoding API v4 converts addresses to coordinates (forward geocoding) and coordinates to addresses (reverse geocoding), and resolves Place IDs. It supports component filtering and region biasing, which matters for India where an address string without the right context can resolve to the wrong state. Lepton integrates it as part of the addressing stack.

Get a demo →

What the API does

Forward geocoding, reverse geocoding and Place ID resolution

The Geocoding API v4 is the current production release (a v3 to v4 migration guide is available; teams on v3 should plan the move). It converts a free-text address to latitude/longitude coordinates (forward geocoding) and converts a latitude/longitude pair back to a human-readable address (reverse geocoding). It also resolves addresses to and from Place IDs for stable cross-API referencing.

Component filtering lets you scope a geocode request to a country, region, locality or postal code, which reduces ambiguous results when address strings alone are underspecified. Region biasing applies a regional preference without a hard filter. JSON and XML response formats are supported. Note: EEA regulatory changes to the Geocoding API are effective from 2025-07-08 for applicable use cases.

Google Maps Platform · Geocoding API v4
GET /v4/geocode/address · a shorthand Bengaluru address, forward geocoded
host geocode.googleapis.com · auth X-Goog-Api-Key · v4 is the current release; a v3 to v4 migration guide is published
200 OK
1 result · ROOFTOP

Request query parameters

GET https://geocode.googleapis.com/v4/geocode/address
  ?addressQuery=42 3rd crs, 100ft rd, indranagr, blr 560038
  &regionCode=IN
  &languageCode=en
X-Goog-Api-Key: <your key>
This is the India problem in one line. The input is shorthand: no house-number convention, an abbreviated road, a misspelt locality and a three-letter city. regionCode keeps the resolution inside India, and granularity tells you how far to trust the pin: ROOFTOP, RANGE_INTERPOLATED, GEOMETRIC_CENTER or APPROXIMATE. Gate on that value rather than on the fact that a result came back at all.

Response application/json

{
  "results": [
    {
      "place": "places/ChIJXXXXXXXXXXXXXXXXXXXXXXX",
      "placeId": "ChIJXXXXXXXXXXXXXXXXXXXXXXX",
      "formattedAddress": "42, 3rd Cross Rd, Indiranagar, Bengaluru, Karnataka 560038, India",
      "location": { "latitude": 12.97142, "longitude": 77.64108 },
      "granularity": "ROOFTOP",
      "types": ["street_address"],
      "addressComponents": [
        { "longText": "42", "shortText": "42",
          "types": ["street_number"], "languageCode": "en" },
        { "longText": "3rd Cross Road", "shortText": "3rd Cross Rd",
          "types": ["route"], "languageCode": "en" },
        { "longText": "Indiranagar", "shortText": "Indiranagar",
          "types": ["sublocality", "sublocality_level_1"], "languageCode": "en" },
        { "longText": "Bengaluru", "shortText": "Bengaluru",
          "types": ["locality"], "languageCode": "en" },
        { "longText": "Karnataka", "shortText": "KA",
          "types": ["administrative_area_level_1"], "languageCode": "en" },
        { "longText": "560038", "shortText": "560038",
          "types": ["postal_code"], "languageCode": "en" },
        { "longText": "India", "shortText": "IN",
          "types": ["country"], "languageCode": "en" }
      ]
    }
  ]
}
results1 object in results[], the array is the responsecomponents7 returned, and formattedAddress is those same 7 in ordergranularityROOFTOP, one of the four documented precision values
Fig. · One forward geocode, trimmed for the page. The endpoint, the query parameters and every response field are the documented Geocoding API v4 surface. The address, Place ID, coordinates and components are illustrative and describe no real address or customer record.

Use cases

Where teams rely on geocoding

  1. 01

    Last mile · Resolving delivery addresses to a pin

    Before a parcel reaches a rider, the address string needs a coordinate. Geocoding API converts customer-entered addresses to pins, with component filtering scoped to the India country context to reduce cross-state mis-resolves on ambiguous locality names.

  2. 02

    Proptech · Plotting and clustering property listings

    Property portals geocode listing addresses to render them on a map, cluster nearby listings and calculate neighborhood distances. Reverse geocoding on a map tap converts the pin back to a readable address for search context.

  3. 03

    Logistics · Batch geocoding for address databases

    Teams with large address datasets use the Geocoding API to resolve historical records to coordinates for route planning, coverage analysis and delivery-zone assignment.

  4. 04

    Place IDs · Stable place references across APIs

    A Place ID is a persistent identifier for a location that works across the Places, Geocoding and Directions surfaces. We wire Place ID resolution so your product references places stably rather than re-geocoding on every request.

Where Lepton comes in

India-biased configuration, v3 migration and the full addressing stack

Geocoding API is rarely the only addressing tool in the stack. It works alongside Address Validation for entry-point verification and Address Descriptors for last-mile landmark resolution. We integrate the right combination and configure each correctly for India.

  1. 01

    Stack · Geocoding as part of the addressing layer

    We design the addressing architecture: where Geocoding handles batch resolution, Address Validation catches errors at entry, and Descriptors add landmark context for Indian pins. One integrated stack rather than three standalone calls.

  2. 02

    Config · Region biasing and component filtering for India

    Ambiguous Indian locality names resolve differently with and without region/component filters. We configure the correct filter strategy so addresses resolve to the right state and city, not a namesake in another region.

  3. 03

    Migration · v3 to v4 structured migration

    Teams still on Geocoding v3 need a planned move. We run the field mapping and response-schema audit, validate against your address corpus and execute the cutover with parallel-run validation.

Get started

Show us the addressing gap. We will close it.

Bring a sample of your address quality problem. We map Geocoding v4, Address Validation and Address Descriptors to where each one moves the number.

Get a demo →