$32 per 1,000 requests. That’s what Google charges for a Text Search call on the Places API. Not for rendering a map. For searching businesses on it.
If you came here trying to decode Google Maps API pricing before it decodes your credit card, you’re in good company. Half the first page of Google for this exact query is forum threads asking someone, anyone, to explain the bill. The short version: Google Maps API pricing is a pay-as-you-go, per-SKU model where every call type, from a map load to a business search, bills at its own rate and carries its own free monthly cap.
Since March 2025, Google Maps Platform has no shared free credit. Each SKU has its own free cap (10,000 calls for Essentials, 5,000 for Pro, 1,000 for Enterprise), then pay-as-you-go rates from $2 to $32 per 1,000 requests. Map rendering is the cheap part. Places data is the expensive part, and it is where alternatives save the most. The FlyByAPIs Google Maps Extractor bills that same business data per request, from $19.99 for 15,000 records with 100 free every month.
$32
Per 1K Text Search calls (priciest SKU)
$200
Monthly credit Google killed in March 2025
10,000
Free monthly calls per Essentials SKU
$0.30
Per 1K requests, Mega plan fully used
We run a suite of data extraction APIs, including a Google Maps business data API that answers millions of requests, so we spend an unhealthy amount of time comparing our invoices to Google’s rate card. This post is that homework, written down.
By the end you’ll know exactly what each SKU costs in 2026, what the free tier really covers, where the three billing traps hide, and which cheaper alternative fits your specific use case. With real math, not “contact sales” hand-waving.
The March 2025 change: your $200 credit is gone
For years, every Google Maps Platform project got a $200 monthly credit that applied across all APIs. Maps loads, geocoding, a few Place Details calls: as long as the blended total stayed under $200, you paid nothing. Most small apps never saw an invoice.
In March 2025 Google replaced that with something structurally different. Every SKU now has its own free usage cap, and every SKU belongs to one of three tiers.
Essentials
10,000 free calls per SKU per month. Basic maps, geocoding, basic Place Details, simple routes.
Pro
5,000 free calls per SKU per month. Text Search, Nearby Search, Place Details with contact data, Street View.
Enterprise
1,000 free calls per SKU per month. Navigation, fleet routing, photorealistic 3D tiles.
Why does this matter so much? Because the caps do not pool. Under the old model, heavy geocoding could borrow headroom from light map usage.
Now each SKU meters independently. A project that comfortably fit under $200 of mixed usage can cross a single Pro cap and start paying $32 per 1,000 requests the same day.
Bottom line:
The March 2025 change didn't raise the rates. It removed the buffer that hid them. The per-call prices were always this high; now you actually meet them.
Google Maps API pricing by SKU: the full 2026 rate table
These are the pay-as-you-go rates per 1,000 requests, verified against Google’s official pricing list in August 2026. Rates drop at 100,000 and again at 500,000 monthly requests.
| SKU | Tier | Free/month | 0-100K | 100K-500K | 500K+ |
|---|---|---|---|---|---|
| Static Maps | Essentials | 10,000 | $2.00 | $1.60 | $1.20 |
| Autocomplete (per request) | Essentials | 10,000 | $2.83 | $2.27 | $1.70 |
| Geocoding | Essentials | 10,000 | $5.00 | $4.00 | $3.00 |
| Place Details (Essentials) | Essentials | 10,000 | $5.00 | $4.00 | $3.00 |
| Compute Routes (Essentials) | Essentials | 10,000 | $5.00 | $4.00 | $3.00 |
| Dynamic Maps | Essentials | 10,000 | $7.00 | $5.60 | $4.20 |
| Compute Routes (Pro) | Pro | 5,000 | $10.00 | $8.00 | $6.00 |
| Dynamic Street View | Pro | 5,000 | $14.00 | $11.20 | $8.40 |
| Place Details (Pro) | Pro | 5,000 | $17.00 | $13.60 | $10.20 |
| Address Validation | Pro | 5,000 | $17.00 | $13.60 | $10.20 |
| Text Search (Pro) | Pro | 5,000 | $32.00 | $25.60 | $19.20 |
| Nearby Search (Pro) | Pro | 5,000 | $32.00 | $25.60 | $19.20 |
| Navigation | Enterprise | 1,000 | $25.00 | $20.00 | $15.00 |
Two things jump out of that table. First, the Maps Embed API and the mobile Maps SDKs are still free with unlimited usage, which is why simple “show a pin on a map” apps often owe nothing. Second, the two most expensive rows aren’t maps at all. They’re searches for business data.
Do you see what this means? If your workload is rendering maps, Google’s pricing is annoying but survivable. If your workload is pulling business listings, you’re buying the most expensive SKUs on the entire rate card.
What you can still get for free in 2026
The People Also Ask box for this keyword is basically four variations of “is it free?”. Fair question. Here’s the honest answer.
You can run a genuinely free setup if your usage fits inside the per-SKU caps. A local app doing 8,000 geocodes, 9,000 map loads, and 4,000 basic Place Details calls a month pays Google exactly $0. Add the unlimited free Embed API and mobile SDKs, and a surprising number of small products stay at zero.
The free tier stops being free at very specific points:
You cross a single Pro cap
5,000 free Text Search calls sounds like a lot. That's about 165 searches a day. One prospecting script burns that in an afternoon, and call 5,001 bills at $0.032.
Your field mask promotes you to a pricier tier
Ask Place Details for a phone number or website and you're no longer on the $5 Essentials SKU. You're on the $17 Pro SKU, with half the free cap.
You forget quotas exist
There is no built-in spending ceiling. A retry loop gone wrong bills at full rate until you notice. Set daily quotas in the Cloud Console on day one.
If you want a quick estimate for your own stack, skip Google’s pricing calculator and do this instead:
- List every SKU your app touches.
- For each SKU, take your monthly calls and subtract its free cap.
- Multiply whatever is left by the 0-100K rate from the table above.
It overestimates slightly at high volume, which is exactly the direction you want an estimate to err.
The real cost: three worked scenarios
Rate tables are abstract. Invoices are not. Here’s what three realistic business-data workloads actually cost, using pay-as-you-go rates and counting the free caps.
Scenario 1: a one-off list of 2,000 leads. Say 100 Text Search queries returning 20 places each. That fits inside the 5,000-call Pro cap, so Google charges $0. Genuinely free, as long as you stay a small fish.
Our Google Maps data extraction API is also free here: 100 requests × up to 200 businesses on the free plan gets you the same 2,000 records many times over for $0. At hobby scale, nobody should be paying anyone.
At hobby scale, both routes are free. The moment volume grows past the free caps, the two bills stop looking anything alike.
Scenario 2: an agency refreshing 50,000 business records monthly. Local SEO reporting, review tracking, that kind of work. You need Place Details Pro for each record: 50,000 calls, minus 5,000 free, is 45,000 × $0.017 = $765/month.
The same 50,000 records through the business details endpoint of the FlyByAPIs Google Maps scraper API run on the Ultra plan at $89.99/month with 150,000 requests included. About 8.5× cheaper, and you have 100,000 requests of headroom left.
Scenario 3: prospecting at scale, 500,000 businesses monthly. With details per business, Google’s math gets ugly: 5,000 free, then 95,000 calls at $17 and 400,000 at the $13.60 volume rate. Roughly $7,000/month, before you even add the search calls that found the businesses.
The extraction route is one request per 200 businesses, with details in the same response. Budget 2,500 search requests, and even allowing generously for duplicates, call it 4,000. That fits the $19.99 Pro plan.
Prefer refreshing every record individually through the business details endpoint? 500,000 requests still only cost $149.99 on Mega.
2,000 records, one-off
$0 vs $0
Both free at this scale
50K records/month
$765 vs $89.99
Google vs FlyByAPIs Ultra
500K records/month
~$7,000 vs $89.99
Google vs FlyByAPIs Ultra
Two honest caveats on that comparison, because I’d rather undersell it than get called out. Dense areas mean duplicate results across searches, so you rarely bank a full 200 unique businesses per request. Budget for 60-70% efficiency.
And Google’s official response surface is broader. If you need live opening-hours guarantees or the full licensed field set inside a consumer-facing map, the official Places API is what you’re paying that premium for.
For internal datasets, lead lists, and market analysis? The premium buys you nothing.
100 requests/month free · No credit card required
Three cost traps that quietly inflate your bill
After reading a depressing number of “why is my bill $900?” threads on Reddit and Stack Overflow while researching this post, the same three traps kept showing up.
Trap 1: the field mask tier promotion. The same Place Details endpoint bills as three different SKUs depending on which fields you request. IDs and address? Essentials, $5. Phone, website, ratings? Pro, $17. Some newer fields? Enterprise.
Most developers pick fields based on what they need and never realize each addition can triple the rate. Audit your field masks before anything else.
The same endpoint, three different prices. Your field mask, not your traffic, decides which rate you pay.
Trap 2: per-SKU caps punish mixed usage. Your app probably touches five or six SKUs. Five of them sit comfortably in free territory while one, usually Autocomplete or a search SKU, quietly crosses its cap and generates the whole invoice. The old $200 credit averaged this out. Nothing averages it out anymore.
Trap 3: Autocomplete bills per keystroke request. At $2.83 per 1,000 requests, a search box that fires on every keystroke turns one user typing “restaurant” into ten billable calls. Use sessions or debounce aggressively. In more than one of those billing threads, autocomplete alone accounted for a third of the invoice.
Pro tip:
Before hunting for alternatives, spend 30 minutes on quotas, field masks, and autocomplete sessions. Teams in those same threads report cutting their bill 40% without changing providers. Then compare what is left against the alternatives below.
Cheaper alternatives, split by what you actually use
This is where most “Google Maps API alternatives” articles go wrong: they pitch one replacement for everything. But Google Maps Platform is really three products wearing one trench coat, and each has different alternatives.
Map tiles and SDKs
You render maps in your product. Alternatives: MapLibre + OpenStreetMap (free), Mapbox, Radar. Or exploit Google's own free unlimited Embed API and mobile SDKs.
Geocoding and routes
You convert addresses to coordinates or compute directions. Alternatives: OSM's Nominatim (free with usage limits), dedicated geocoders at $0.50-2.00 per 1K.
Places and business data
You need listings, phones, websites, ratings, reviews. This is the $17-32/1K territory, and where extraction APIs run 8× to 78× cheaper in the scenarios above.
Honest disclosure
We compete in exactly one of these quadrants: business data. For tiles and navigation, we will happily tell you to use someone else.
For map rendering, the free path is real: MapLibre rendering OpenStreetMap data powers plenty of production apps at $0 for the software. You pay only for tile hosting or use a hosted provider. If your entire Google bill is Dynamic Maps loads, you can probably eliminate it.
For geocoding, Google’s $5.00 per 1,000 is mid-pack. Nominatim is free for light usage under its fair-use policy, and several dedicated geocoding services undercut Google by 60-90% at volume. Not our fight, so I’ll leave it there.
For places and business data, the gap is the widest, and it is structural. Google prices place data as a premium add-on to a maps product. An extraction API prices it as the whole product. That is the segment the rest of this post digs into, because it is where a Google Maps extraction API replaces the two most expensive SKUs on the rate card.
Places API vs an extraction API: the lead-gen math
Let me make the comparison concrete with code, because “cheaper” claims without a request you can copy-paste are just marketing.
Getting 20 businesses with full contact data out of the official Places API is a two-step dance: one Text Search Pro call ($0.032) to find them, then twenty Place Details Pro calls ($0.017 each) to get phones and websites. Total: $0.372 for 20 businesses, or about $18.60 per 1,000 records once you’re past the free caps.
Doing the same through a per-request Google Maps API takes one request. Search and details come back together:
| |
Every record in data already includes the name, full address, phone, website, rating, review count, categories, coordinates, and opening hours. No second call, no field mask tier roulette. The locate_and_search endpoint docs
list the full response schema. Pagination runs through offset, 20 results at a time.
And here’s the plan-level comparison, using what you’d actually pay on the Google Maps Extractor plans rather than cherry-picked marginal rates:
| Monthly need | Google Places (pay-as-you-go) | FlyByAPIs plan | You pay |
|---|---|---|---|
| Up to 2,000 records | $0 (inside free caps) | Basic (free), 100 requests | $0 either way |
| ~50,000 records with details | ~$765 | Ultra, 150K requests | $89.99 |
| ~500,000 records with details | ~$7,000 | Ultra via search, Mega for 1:1 refresh | $89.99-$149.99 |
| Rate limit | Custom quotas | 4-15 requests/second by plan | Included in plan price |
One more thing worth knowing if you compare us against other extractors rather than Google: every request is routed through an IP inside the target country. Searching German businesses? The request comes from a German IP, so prices, rankings, and review counts match what a local actually sees. Cross-country inconsistency is the most common data-quality complaint we hear from people migrating off other scrapers.
And if maps data is one piece of a bigger pipeline, the same per-request model runs across our whole suite: check a prospect’s visibility with the Google Search ranking API , pull funding data through the Crunchbase company data API , spot hiring signals via the job listings API , track their products with the Amazon product data API , and normalize multilingual reviews through the translation API . One billing model, no SKU matrix.
How to decide in five minutes
Stay on Google
You embed consumer-facing maps with live data, need the licensed field set, or fit inside the free caps. Set quotas, audit field masks, and the bill stays sane.
Go open source
Your bill is mostly map loads and light geocoding. MapLibre plus OpenStreetMap kills that line item entirely, at the cost of some setup work.
Switch to extraction
You pull business data for lead gen, market research, or monitoring. The Pro SKUs are your bill, and a per-request extractor cuts it by 8× to 78×.
Mix both
Plenty of teams keep Google for the consumer-facing map and move bulk data collection to an extractor. The two are not mutually exclusive.
The part nobody puts in the pricing table
I’ll close with the thing that actually surprised me writing this. Google’s rates did not change in March 2025. The $32 Text Search rate existed for years. What changed is that the credit hiding it disappeared, and thousands of teams met their real cost structure for the first time.
That’s worth sitting with. The best time to understand your per-request economics is before the invoice teaches you. Whether that leads you to field-mask audits, MapLibre, or a flat-rate Google Maps business data API , you now have the actual numbers to decide with.
If your workload is the business-data kind, the free tier is the honest way to test the claim: pull a few hundred records, compare the fields against what you get from Places, and check the math against your own volume.
100 requests/month free · No credit card required
P.S. If you also run search workloads, we did the same exercise for SERP data: the Google Search API cost breakdown follows the exact same “the credit was hiding the rate” pattern. Google is nothing if not consistent.
Oriol.
