| SKU | $ / 1,000 | Calls / month | Cost |
|---|
Why per-SKU is the only honest way to budget Maps
Google Maps Platform isn't one price — it's dozens. Loading an interactive map, geocoding an address, drawing a route and computing a distance matrix are four completely different SKUs at four different prices, and the spread is large: a Dynamic Map load and a Directions Advanced call can differ by more than 3×. That means the single most useful thing you can know before launch isn't your total call count, it's which SKU you lean on hardest. An app that geocodes once and shows a static map is cheap; an app that recomputes a live Distance Matrix on every search can quietly run into four figures a month at the same user count.
This calculator lays out the major SKUs side by side, takes your monthly call volume for each, and applies a configurable free credit so you see both the gross cost and what you'd actually pay. The "biggest SKU" readout points straight at the line worth optimising — usually by caching geocodes, batching matrix calls, switching a live map to a static image where interaction isn't needed, or moving a heavy route feature behind a user action instead of running it on page load. Once you know the Maps number, fold it into your whole infrastructure bill with the API stack cost calculator, and if Maps is dominating, compare alternatives in the maps API guides and alternatives finder.
The SKUs that surprise people
Distance Matrix bills per element. A request with 10 origins and 10 destinations is 100 billable elements — multiply, don't count requests. Autocomplete can bill per keystroke. Without session tokens, every character a user types can be a separate billable request; with sessions, the whole lookup bills once. Dynamic vs static maps. An interactive map costs per load every time the page mounts; a static map image is far cheaper if users don't need to pan and zoom. Places Details fields matter. Asking for more field categories can move the call into a higher-priced SKU. Modelling these correctly is the difference between a believable budget and a shock.
How to use it
1. Enter your expected monthly calls for each SKU you use (leave others at 0).
2. For Distance Matrix, enter elements (origins × destinations), not requests.
3. Set the monthly free credit to whatever your account actually gets — or 0 to see the full gross cost.
4. Read the net bill and the biggest SKU, then optimise that line first.
Common mistakes
Counting matrix requests instead of elements. The per-element unit is the classic Maps bill shock. Forgetting Autocomplete session tokens. Per-keystroke billing can be 5–10× the session cost. Loading dynamic maps where static would do. Pay per interactive load when an image would have been fine. Assuming the free credit covers you forever. It's a fixed monthly amount; growth blows past it. Ignoring volume discounts. At scale Google's price per call drops in tiers, so high-volume estimates here are conservative.
FAQ
How does Maps billing actually work?
Per SKU: each product has its own price per 1,000 calls. All SKU usage is summed, your free credit is subtracted, and you pay the rest. Different SKUs cost very differently.
What is one billable call?
Depends on the SKU — per map load, per request, or per element. Distance Matrix is per element (origins × destinations), which is the most common source of surprise.
Is there still a $200 free credit?
Google has shifted toward per-product free allowances and the structure varies by account. Set the free-credit field to your real recurring amount, or 0 for gross cost, and confirm in the console.
How do I cut a high Maps bill?
Cache geocodes, use session tokens for Autocomplete, batch or limit Distance Matrix elements, and swap dynamic maps for static images where interaction isn't needed.
Estimate only. SKU prices and free-credit terms are reference figures and change — verify in the Google Cloud console.