pytrends alternative: Google Trends API for Python, with 429 retries handled remotely
pytrends is archived and Google Trends blocks it with 429 Too Many Requests. Google's official Trends API is still in closed alpha. This small library gives you the same TrendReq interface as pytrends, but the requests run on a managed scraper: each request is paced and retried with a fresh Google session; if Google keeps refusing, it falls back to the embeddable chart pages and, last, to a real browser. Retries and time limits are fixed, so runs end instead of hanging. Google still refuses some requests. Terms that still fail are not billed.
from gtrends_api import TrendReq # was: from pytrends.request import TrendReq
pytrends = TrendReq() # reads APIFY_TOKEN from the environment
pytrends.build_payload(["coffee", "tea"], timeframe="today 12-m", geo="US")
df = pytrends.interest_over_time()
print(df.tail())
Output of a real run on October 10, 2026 (about 30 seconds):
coffee tea isPartial
date
2026-09-06 74 31 False
2026-09-13 74 31 False
2026-09-20 74 33 False
2026-09-27 70 31 False
2026-10-04 70 30 True
Is there an official Google Trends API?
Yes, in closed alpha: Google opened applications on July 24, 2025 and, as of October 10, 2026, still admits "a limited set of developers" (API page). If you're not in it, this library gives you the pytrends interface today. Other official options: the CSV download on the Google Trends site, the Google Trends BigQuery dataset and the Trending Now RSS feed.
Why switch
| pytrends | this library |
|---|
429 Too Many Requests | Frequent, especially from servers and notebooks | Still happens; paced, retried with a new Google session, then the embeddable charts and a real browser |
| Maintained | Archived (read-only) since April 2025 | Maintained |
| Interface | TrendReq, build_payload, interest_over_time… | Same method names, same DataFrame shapes |
| Hundreds of keywords | Manual batching, not comparable between batches | Available via raw() with an anchor term on one scale |
| Cost | Free, but blocked | Pay per keyword: $0.003 per keyword (about $3 per 1,000) on Apify's Free and Starter plans, less on Scale ($2.20 per 1,000) and Business ($1.50 per 1,000). Each method call is billed separately (see Pricing). Apify's free plan includes a small monthly credit, enough to try it |
Install
pip install git+https://github.com/Egrarobo/pytrends-alternative.git
Get an API token: create a free account at apify.com, then copy the token from Settings → API & Integrations.
export APIFY_TOKEN="your-token"
Supported methods
| pytrends method | Supported | Notes |
|---|
build_payload(kw_list, cat, timeframe, geo, gprop) | Yes | 1–5 terms, like trends.google.com |
interest_over_time() | Yes | DataFrame indexed by date, one column per term, plus isPartial |
interest_by_region(resolution, inc_low_vol, inc_geo_code) | Yes | COUNTRY, REGION, CITY, DMA |
related_queries() | Yes | {term: {"top": DataFrame, "rising": DataFrame}} |
related_topics() | Yes | Same shape, with title and topicType |
trending_searches(pn) | Yes | pn="united_states" or an ISO code like "GB" |
raw() | New | Full JSON per term: averages, peak date, top region, explore link |
Timeframes: now 1-H, now 4-H, now 1-d, now 7-d, today 1-m, today 3-m, today 12-m, today 5-y, all, or a date range such as 2025-01-01 2025-12-31.
More examples
Top US states for a keyword
pytrends.build_payload(["solar panels"], timeframe="today 12-m", geo="US")
states = pytrends.interest_by_region(resolution="REGION", inc_geo_code=True)
print(states.sort_values("solar panels", ascending=False).head(10))
Rising related searches (content ideas)
pytrends.build_payload(["air fryer"], timeframe="today 3-m", geo="GB")
rising = pytrends.related_queries()["air fryer"]["rising"]
print(rising[["query", "formattedValue"]])
Today's trending searches
print(pytrends.trending_searches(pn="united_kingdom")[["rank", "title", "approxTraffic"]])
Many keywords on one scale (more than 5 terms, comparable with each other): call the scraper directly with an anchor term, see the scraper page.
Pricing
You pay Apify per keyword for which all requested data was collected:
| Apify plan | Price per keyword | Per 1,000 keywords |
|---|
| Free, Starter | $0.003 | $3.00 |
| Scale | $0.0022 | $2.20 |
| Business and higher | $0.0015 | $1.50 |
Plus a tiny start fee per run ($0.00005). Failed keywords are free.
Each method call is a separate run on Apify and is billed per keyword. For example, 2 keywords with interest_over_time(), interest_by_region() and related_queries() are billed as 6 keywords. To get the timeline, regions and related queries in one run, billed once per keyword, use:
data = pytrends.raw(interestOverTime=True, interestByRegion=True, relatedQueries=True)
raw() returns the full JSON instead of DataFrames. Repeating the same method call with the same payload is cached and not billed again.
No Python? Use it from anywhere
The same data is available without code: run the Google Trends Scraper in the browser, schedule it, export to Excel or Google Sheets, or call it from Make, Zapier, n8n or any HTTP client.
Notes
- Values are Google's relative 0–100 index, not search counts. That is how Google Trends works.
- You pay only for terms where all requested data was collected; failed terms are free.
- This project is not affiliated with Google or with the pytrends authors.
License
MIT
