keyword source · pypi

PyPI Trends API

Weekly download counts for any PyPI project as a normalized time series, with growth over 3M/6M/12M/5Y already computed. Python adoption signals without loading pypistats or BigQuery dumps.

POST get_growth · python

# one POST, every source
POST https://api.trendsapi.ai/api
Authorization: Bearer YOUR_API_KEY

{ "mode": "get_growth",
  "source": "python",
  "keyword": "fastapi",
  "window": ["3M", "12M"] }

200 OKnormalized 0-100

response

  {
    "search_term": "fastapi",
    "data_source": "python",
    "results": [
      {
        "period": "3M",
        "growth": 11.8,
        "direction": "increase"
      },
      {
        "period": "12M",
        "growth": 92.4,
        "direction": "increase"
      }
    ]
  }

application/jsontimestamped

Why it exists

PyPI trend data, without the plumbing

PyPI download data is public, but using it means pypistats for quick checks or the public BigQuery dataset for anything serious, and both stop at raw counts. Trends API normalizes those counts to a 0-100 series, computes growth automatically, and returns the same response shape as npm, GitHub and every other source.

That matters most in the AI and data ecosystem, where Python is the default language. Watching install curves for frameworks, vector clients and orchestration libraries is an early read on which tools developers actually adopt, not just talk about.

Use cases

What people build with it

Tech selection

Check whether a library's install curve is still climbing before betting a codebase on it.

AI and data teams

Track the ML stack: frameworks, vector clients and orchestration libraries on one scale.

DevRel

Measure install growth after each release, talk or docs push.

Dev-tool investing

Watch ecosystem momentum across competing Python packages.

Comparison

How it compares

	pypistats / BigQuery	Trends API
Data returned	raw install counts	counts plus 0-100 score
Normalization	DIY	0-100, done
Growth rates	compute yourself	3M/6M/12M/5Y built in
Compare vs npm or search	no	same scale, one endpoint
Free tier	free	100 requests/month

Three modes, one endpoint

get_time_series

Historical trend series for a keyword on this source, normalized 0-100.

get_growth

Growth percentages over 3M / 6M / 12M / 5Y windows.

get_top_trends

24 live trending feeds across platforms, no keyword needed.

FAQ

Common questions

What PyPI data does Trends API provide?

Weekly download counts for any PyPI project as a normalized time series, plus growth percentages over 3M/6M/12M/5Y windows.

Why is the source string python and not pypi?

python is the source value the API expects. The keyword is still the exact PyPI project name in its normalized registry form: pandas works, Pandas does not.

How is this different from pypistats or BigQuery?

Those return raw install counts for date ranges you specify. Trends API normalizes to 0-100, computes growth automatically, and returns the same response shape as npm, GitHub and every other source.

Are conda installs included?

No. The series is PyPI downloads. Conda, Homebrew and system packages are separate channels.

Can I compare a Python package with an npm package?

Yes. One call per package, then compare the normalized scores. The 0-100 rescaling makes cross-ecosystem comparison meaningful even when absolute counts differ by 100x.

Keep exploring

Related trend sources

npmnpm trend data as JSON GitHubGitHub trending repositories as JSON Google TrendsQuery Google Trends data as JSON

Start pulling PyPI trends in 60 seconds.

100 requests a month free. No credit card, no sales call. One key works for every source.