Google Reviews Batch API for Fast Date-Range Exports

The Google Reviews batch API is now available to SERP Search customers on Pro, Business, and Enterprise plans. Give it a Google Maps location and a creation-date range, and it produces a downloadable review dataset as an asynchronous job. Your application gets a job ID immediately, can check progress without holding a connection open, and can download a private gzipped NDJSON file when the export is ready.

This endpoint is built for historical exports, recurring data collection, review analysis, and other jobs where requesting one review page at a time would add avoidable client code and waiting. It also has a delivery-based usage policy: a failed or cancelled export doesn't consume monthly API requests.

What the Google Reviews batch API supports

Each export targets one Google Maps location. You can identify that location with either a Google Place ID or a Feature ID, then choose an exact review creation-time window.

Capability	What to expect
Date range	RFC 3339 startAt and endAt values. The start is included and the end is excluded.
Location IDs	A Google Place ID or Feature ID can identify the location.
Locale	Optional hl and gl values select the language and country context.
Review filter	An optional search value filters review text. Omit it to export every review Google returns in the range.
Job status	Jobs report queued, running, completed, failed, or cancelled.
Progress	Status responses include page, matched-review, and exported-review counts.
File format	Gzipped NDJSON with one review object per line, using the same review schema as the standard endpoint.
Access	Downloads are private and owner-scoped to the API account that created the job.

The job endpoints cover creation, listing, status checks, cancellation, and download access. An optional Idempotency-Key also makes it safe to retry a creation request without accidentally starting the same export twice. The review batch API reference lists every request field, response field, status, and error code.

Sequential reviews compared with batch exports

The useful comparison is between the standard Google Maps Reviews endpoint and a batch export. Both use the same review schema, but they handle pagination, delivery, and failure differently.

Capability	Sequential Reviews endpoint	Batch Reviews endpoint
Request flow	Request one page, read next_page, then request the following page.	Create one asynchronous job, check its status, then download a file.
Client runtime	Your client coordinates the full cursor chain.	Your client can disconnect while the export continues.
Output	One JSON response per page.	One private gzipped NDJSON file for the selected date range.
Recovery	Your application stores progress and resumes interrupted crawls.	The service tracks job progress and preserves the finished artifact.
Usage accounting	Each Reviews API request consumes quota as it succeeds.	Page requests are charged together after an artifact is delivered.
Best fit	One page, a few pages, or latency-sensitive interactive requests.	Historical ranges, complete datasets, archives, and offline analysis.

Sequential pagination is a dependent chain. The request for page two cannot begin until page one returns its cursor, and the same dependency continues through the final page. A batch request returns a job immediately and moves that long-running collection out of your application's request path.

That does not mean every batch job is guaranteed to finish faster than every sequential crawl. The standard endpoint is usually the faster choice when you only need one page. The batch advantage becomes relevant as the requested history grows because your client no longer has to remain active, assemble page responses, persist cursors, or recover a half-finished export.

How to compare completion time

A valid speed benchmark must use the same location, locale, date range, and final review-ID set on both paths. Sequential time runs from the first Reviews request through the final required page. Batch time runs from job creation through availability of the downloadable artifact.

We don't publish a fixed speed multiplier because the result changes with the number and distribution of reviews, current service load, and Google's response time. The batch API saves client orchestration time by design, while end-to-end wall-clock savings depend on the export.

How API usage and cost are calculated

Batch review usage is based on Google review pages requested, not the number of reviews returned and not the number of jobs used to complete an export. One page request equals one request from your monthly API quota.

If an export takes 931 page requests and delivers its artifact, requestsCharged will be 931. A review entry itself isn't a separate usage unit, so an export with thousands of reviews doesn't automatically cost thousands of requests. The page count needed depends on the location, the selected date range, and what Google returns during the job.

Do not estimate a batch charge from review count or from the number of pages used by a separate sequential crawl. A batch export can use a different page total for the same location and date range. The job's final requestsCharged value is the authoritative cost, and repeated exports of the same dataset can use a slightly different number of requests.

When usage is charged

Usage is reserved while an export is running. Reserved requests reduce the quota available to other calls, but they aren't counted as consumed usage yet. The final page total is charged once the API has a downloadable artifact to deliver.

Job outcome	Monthly quota result
Full export delivered	Charge the final number of page requests.
Quota-limited partial export delivered	Charge the page requests used before the quota limit was reached.
Export fails without a delivered file	Charge zero requests and release the reservation.
Export is cancelled	Charge zero requests and release the reservation.

Creating, listing, polling, cancelling, and requesting a download link don't deduct monthly request quota. Those calls still count against the account's requests-per-second limit.

The reservation is visible through X-Quota-Reserved, while requestsCharged on the job shows the final charge. For current plan quotas and top-up prices, check the SERP Search pricing page.

What happens when quota runs out

A large export won't silently exceed the account's remaining monthly quota. If the job reaches that exact limit before covering the full date range, it stops and delivers the reviews collected so far. The completed job reports truncatedByQuota: true and charges the available requests.

For example, if an account has 900 requests left and an export needs more than 900, the delivered partial file is charged 900 requests and the account's remaining quota becomes zero. If no artifact can be delivered because the job fails, those reserved requests are released instead.

Applications should record both requestsCharged and truncatedByQuota. Together, those fields tell you the cost of the file and whether it covers the full requested range.

What speed should you expect?

Production completion time varies. A location with more available review pages can take longer. Date-range density, locale, text filtering, current service load, and Google's response time can also affect a run. Treat batch completion as asynchronous processing rather than a fixed-latency request.

The API is asynchronous for that reason. POST /api/v1/reviews/batches returns a job instead of making your client wait for the final file. You can poll GET /api/v1/reviews/batches/{id} at a cadence that fits your application, then request a download URL after the status becomes completed.

Completed files remain available for seven days. Each signed download URL is valid for 15 minutes, and you can request a fresh URL during the retention period. An account can have up to three queued or running review exports at once.

For latency-sensitive requests where you only need one page of reviews, the standard Google Maps Reviews endpoint is still the simpler choice. The batch endpoint fits defined date ranges, complete historical datasets, periodic archives, migrations, and offline analysis.

Availability and getting started

Review batch exports are included with active Pro, Business, and Enterprise subscriptions. Free and Starter accounts can continue to use the standard Reviews endpoint, but they can't create asynchronous batch exports.

To run your first export:

1. Find the location's Place ID or Feature ID. If you need to translate between formats, use our Google Maps ID conversion guide.
2. Choose an inclusive UTC start time and an exclusive UTC end time.
3. Create the job with POST /api/v1/reviews/batches and save its ID.
4. Poll the status resource until the job reaches a terminal state.
5. Download completed data within the seven-day retention window.

Read the complete review batch documentation for request examples and schemas. If you need access, compare the eligible plans on pricing before starting a large export.