What the Google News API does and who it is for
The Google News API wraps Google News content into a set of REST endpoints that return categorized news articles, local headlines, and keyword-based search results. Rather than building your own scraper or managing a patchwork of RSS feeds, you send a GET request with a language-region code and get back structured news data in return.
The API suits a fairly broad set of developers: those building news aggregators, content dashboards, research tools, or any application where surfacing current events adds value. The keyword search endpoint also makes it practical for monitoring tools — tracking how a brand, topic, or term is covered across news sources in near real-time.
Endpoint walkthrough
The API exposes twelve distinct endpoints under the base URL https://google-news13.p.rapidapi.com. The core design is straightforward: most endpoints accept an lr parameter (language-region, e.g., en-US) and return articles filtered to that locale.
Category endpoints
Seven endpoints cover fixed editorial categories:
- GET /latest — general top stories across all categories
- GET /world — international news
- GET /business — business and finance coverage
- GET /entertainment — entertainment headlines
- GET /health — health and medicine news
- GET /science — science coverage
- GET /sport — sports news
- GET /technology — technology reporting
For all category endpoints except /world, the lr parameter is required. On /world, it is optional and falls back to a default. This is a small but important distinction: if you forget lr on /latest or /business, for example, the request will fail rather than silently returning a default locale.
Search and discovery endpoints
GET /search accepts a required keyword parameter plus an optional lr. This is the most flexible endpoint — any news topic can be queried rather than being limited to the predefined categories.
GET /search/suggest (also surfaced as /suggest) powers autocomplete. It takes a required keyword representing a partial query string and returns search suggestions, which is useful if you are building a search interface and want to guide users toward well-formed queries.
Local and utility endpoints
GET /local returns local news for a specified region using a gl parameter (e.g., gl=id for Indonesia). This differs from the lr language-region pattern used by most other endpoints — it identifies the country of local interest rather than a combined language-locale string.
GET /languageRegions returns the list of supported language and region codes. Checking this endpoint during setup is worthwhile since passing an unsupported lr value to other endpoints could cause unexpected results.
Pricing breakdown
The API uses a freemium model with four tiers:
| Plan | Price | Monthly requests | Rate limit |
|---|---|---|---|
| BASIC | $0 / month | 25 | 1 req/sec |
| PRO | $25 / month | 10,000 | 5 req/sec |
| ULTRA | $50 / month | 25,000 | 10 req/sec |
| MEGA | $100 / month | Unlimited | 10 req/sec |
The BASIC tier is effectively a trial: 25 requests per month is enough to validate your integration and test response shapes, but it cannot support any production workload. If you are polling even a single category endpoint every hour, you will exhaust the free quota in about a day.
PRO at $25/month offers 10,000 requests, which translates to roughly 13 requests per hour around the clock. That covers a modest production app — a single-user dashboard refreshing several categories a few times per hour would fit comfortably. The rate limit of 5 req/sec also allows small bursts for initial page loads.
ULTRA (marked as recommended) triples the request ceiling to 25,000/month at the same 10 req/sec cap. For applications with meaningful user bases or those polling multiple categories on a schedule, ULTRA is the likely target tier.
MEGA removes the monthly request ceiling entirely while keeping the same 10 req/sec rate limit. This is the appropriate choice for high-throughput applications — for example, an aggregator that continuously monitors many keyword searches or refreshes dozens of locale-specific feeds.
Practical use cases
News aggregators and widgets are the most direct fit. The category endpoints map cleanly onto the sections of a typical news application — world, business, technology, sports — and the lr parameter lets you localize content per user.
Brand and topic monitoring works well through /search. You can poll a keyword on a schedule and store results to track how news coverage of a company, product, or event evolves over time. At the PRO or ULTRA tier, polling several keywords every 30 minutes is feasible within the monthly quota.
Search interfaces benefit from /search/suggest, which adds autocomplete behavior without requiring a separately managed suggestion dataset.
Localized applications can combine /local with the gl parameter to surface hyperlocal headlines alongside global content from the category endpoints.
Performance and reliability
The API reports an average latency of 593 ms and a 99% success rate. A 593 ms average is relatively slow for a synchronous user-facing call — rendering news on initial page load with a direct API call in the critical path would be noticeable to users. Caching responses on your server and refreshing them on a background schedule is the standard pattern here: fetch new articles every few minutes, serve cached data to end users, and avoid per-user API calls altogether.
Limitations and things to check before integrating
- Thin free tier. 25 requests per month on BASIC makes real testing awkward. Budget for at least one month of PRO while building and validating your integration.
- Required
lron most endpoints. A missing language-region parameter will error on six of the eight category endpoints. Build yourlrvalue into a shared request configuration rather than passing it per-call. - Rate limits are shared across all endpoints. The per-second cap applies to your entire usage, not per endpoint. If your application fans out across several categories simultaneously, a burst of parallel requests could hit the limit even at the ULTRA or MEGA tier.
- Response schema. The documentation does not publish a formal response schema. Test the actual JSON structure of each endpoint you plan to use before building your data models around assumed field names.
- Upstream dependency. Because this API surfaces Google News content, any changes Google makes to its underlying feed could affect availability or response structure.
Getting started
Access is via RapidAPI. After subscribing to a plan, include your RapidAPI key in the X-RapidAPI-Key request header on every call. A minimal first request to validate your setup:
GET https://google-news13.p.rapidapi.com/world?lr=en-US
Headers:
X-RapidAPI-Key: YOUR_KEY
X-RapidAPI-Host: google-news13.p.rapidapi.comIf you are unsure which lr values are accepted, call /languageRegions first to retrieve the full list before building out your locale handling logic.
