Google Maps API pricing, explained in 2026
The March 2025 restructure killed the $200 credit. Per-SKU free tiers, three SKU categories, and Nov 2025 subscription plans — plain-English breakdown with worked cost examples at 100k, 500k, and 1M map loads.
If your team's mental model of Google Maps Platform pricing still includes the flat $200 monthly credit, you're budgeting with a number that no longer exists. Google removed the credit on 1 March 2025 and replaced it with per-SKU free thresholds that vary wildly across the catalogue. The November 2025 subscription plans added a second axis — flat-rate monthly plans with predictable bills. This post is the cheat sheet we built internally so we could quote real numbers to prospects.
All numbers below are pulled from the official pricing page, the India pricing page, and the core services pricing list.
The shape of the new model
The March 2025 restructure did three things at once:
- Per-SKU free tier. Every SKU now has its own monthly free allowance. The allowance resets on the first of each month and does not roll over. You must enable a billing account even to claim the free tier.
- Three SKU categories. Products are grouped by complexity and target customer. Each category has a different free allowance:
- Essentials (Maps JavaScript API, Maps Static API, Geocoding, Autocomplete, Places UI Kit, Static Maps, Map Tiles API) — 10,000 free calls/SKU/month. Map Tiles API is the exception at 100,000 free/SKU/mo.
- Pro (Places API Place Details, Text Search, Nearby Search Pro, Places Aggregate, Address Validation Pro, Routes Advanced) — 5,000 free calls/SKU/month.
- Enterprise (Places Insights, Premium SKUs) — 1,000 free calls/SKU/month.
- Auto-volume discounts that scale to 5M+ monthly events. Previously the discount tiers topped out around 100K monthly events.
The mental shift: you no longer budget against a single $200 bucket. You budget against each individual SKU you call, and the right combination depends on your product surface area.
The India wedge — 70k free per Essentials SKU
If your billing address is in India and the majority of usage is in India, the free thresholds jump significantly:
| Category | Global free | India free | Multiplier |
|---|---|---|---|
| Essentials | 10,000 | 70,000 | 7× |
| Pro | 5,000 | 35,000 | 7× |
| Enterprise | 1,000 | 7,000 | 7× |
The 7× bump is the single most important number in this whole post if you're shipping from India. A side project that would burn through the global free tier at 8k monthly map loads sails past India's 70k free ceiling with the same code, unchanged.
Worked cost examples — 100k, 500k, 1M monthly map loads
The pricing-page example below uses Geocoding rates because they are the most granular publicly. Map-load rates are similar in shape but use the Map Tiles API volume tiers.
Assumption: global billing address (not India). 1 Geocoding SKU call per map load.
| Monthly map loads | Free tier used | Paid volume | Effective cost (Tier 1: $5/1K) |
|---|---|---|---|
| 8,000 | 8,000 (within free 10k) | 0 | $0 |
| 100,000 | 10,000 (full free tier) | 90,000 | $450 (90 × $5) |
| 500,000 | 10,000 | 490,000 | $1,960 (90 × $5 + 400 × $4) |
| 1,000,000 | 10,000 | 990,000 | $3,160 (90 × $5 + 400 × $4 + 500 × $3) |
| 5,000,000 | 10,000 | 4,990,000 | $8,110 + high-volume tier savings |
The PAYG schedule for Geocoding is Tier 1 (0–100K) $5.00/1K, Tier 2 (100K–500K) $4.00/1K, Tier 3 (500K–5M) $3.00/1K, Tier 4 (5M+) $1.50/1K, Tier 5 (high volume) $0.38/1K. The volume discount kicks in across the entire billable volume once you cross each tier — at 500k map loads you pay Tier 1 for the first 90k and Tier 2 for the remaining 400k, not Tier 2 for all 490k.
Subscription plans (November 2025)
For predictable monthly bills, Google now sells three subscription plans that pool usage across "combined calls":
| Plan | Monthly price | Included combined calls | APIs covered |
|---|---|---|---|
| Starter | $100/mo | 50,000 | Maps + Geocoding only |
| Essentials | $275/mo | 100,000 | Maps, Routes, Places, Environment |
| Pro | $1,200/mo | 250,000 | All major APIs |
"Combined calls" is a fuzzy unit — a single Map load is one call, but a single Places detail lookup is also one call, even though the underlying cost differs. Subscription plans work when you have predictable, mostly-Maps traffic. They break when your usage skews heavily to Pro-category APIs (which consume calls faster but at a higher unit cost).
The breakeven math:
- Starter $100/50k = $2.00 per 1k calls. Beats PAYG Tier 1 ($5/1K) by 60%, beats Tier 2 ($4/1K) by 50%.
- Essentials $275/100k = $2.75 per 1k. Beats PAYG Tier 1, loses to Tier 2 by 31%.
- Pro $1,200/250k = $4.80 per 1k. Loses to PAYG Tier 2+ unless you also get flat bill predictability.
If your call volume is steady and you want a fixed line item on your SaaS invoice, subscription plans are the right answer. If your traffic is bursty (event launches, marketing campaigns), PAYG with its volume discounts wins.
When Google is still the right choice
Three scenarios where Google Maps Platform remains the pragmatic choice despite the pricing changes:
- You need Places API quality at scale. Google's POI dataset is still the deepest in the industry. If your product depends on rich POI metadata, hours of operation, reviews aggregation, or address validation accuracy, Pro and Enterprise SKUs justify the cost.
- You ship globally and need one vendor. Google covers 200+ countries with consistent data quality. If you can't accept "good in India, mediocre in Brazil," Google is the lowest-friction answer.
- You're already paying > $1,000/mo. At that scale, the subscription plan discount and committed-use discounts outweigh the architectural benefits of switching.
When to look at alternatives
Three scenarios where alternatives win on cost without giving up capability:
- You're under 100k monthly map loads. Pay-as-you-go with a free Mapbox tier (50k map loads) or maps.guru's Pro tier (1M tiles for $49) covers your entire bill with predictable pricing and no per-SKU bookkeeping.
- You're India-only and don't need global POI metadata. Ola Maps (500k free/mo, India-only) or Mappls (India + neighbours) cover the use case at materially lower cost.
- You're already on MapLibre for rendering. Paying Google only for the Places lookup while serving tiles from a non-Google stack gives you Google's POI quality without the rendering premium.
For a side-by-side pricing comparison across the nine vendors we track, see the Google Maps alternatives comparison. For a worked calculator that takes your monthly volume and tells you the cheapest vendor, see the cost calculator.
FAQ
What happened to the $200 Google Maps credit?
Google removed the $200 monthly credit on 1 March 2025. It was replaced by per-SKU free thresholds (10k Essentials, 5k Pro, 1k Enterprise globally; 70k/35k/7k for India-billed accounts). You must enable a billing account to claim the free tier.
Is the India 70k free tier still active?
Yes, as of July 2026. The 7× India multiplier remains on the India pricing page. Eligibility requires an India billing address and majority usage in India.
Do subscription plans replace PAYG?
No, both run in parallel. PAYG charges per call with volume discounts. Subscription plans pool usage across "combined calls" at a flat monthly rate. Most teams run PAYG until they hit a steady $500+/mo bill, then evaluate a subscription plan breakeven.
Does the Maps JavaScript API count as a Map load?
Every Map instantiation on the client counts as one Map load. Pan, zoom, and tile fetches within an existing map do not count as new loads. A single-page web app with one persistent map = one Map load per session.
What counts as a "combined call" in subscription plans?
Subscription plans use a normalised "call" unit that bundles different SKUs into a single counter. The exact mapping is opaque, but in practice Essentials subscription plans cover Maps, Routes, Places, and Environment under the same 100k combined budget.
Can I mix Maps Platform with OpenStreetMap or MapLibre?
Yes. Many teams serve tiles from Mapbox or maps.guru while paying Google only for the Places API lookups. This pattern — Google for POI data, MapLibre for rendering — is the cheapest way to get Google's POI quality.