Wikipedia Trends API

What Wikipedia Trends API MCP server does

Wikipedia Trends API MCP server connects agent requests to the Trends API for Wikipedia page-view data. Its supported operations correspond to three main API methods: get_time_series returns dated values for a keyword, get_growth compares activity across one or more periods, and get_live retrieves a current ranked feed of trending pages. The API also exposes get_top_trends, which uses the same top-trends response type.

Historical results use a 0–100 index for the requested series. Each time-series point includes a date, value, and echoed keyword; volume and pipeline labels may also be present. Growth responses contain one result per requested window, including the period, growth value, direction, dates, and source values. Live results include a snapshot timestamp and ranked [rank, label] rows.

How it works

Requests are sent to POST https://api.trendsapi.ai/api with a bearer token in the Authorization header. Wikipedia history requests use the lowercase source value wikipedia and either the get_time_series or get_growth mode. Live Wikipedia requests use the exact type Wikipedia Trending rather than a source field. Mixing these conventions can produce a 400 response.

A keyword identifies the series to retrieve, so article naming matters. For example, Java and Java (programming language) represent different series. Growth can use presets such as 7D, 3M, 12M, YTD, or 2Y, and it also accepts custom recent and baseline dates. Multiple growth windows are counted as one request. Live results support limit and offset; the documented category option applies to other feed types listed by the API.

Setup and configuration

Install the Python package with pip install trendsapi-wikipedia, then set TRENDSAPI_KEY to a key obtained from trendsapi.ai. Python 3.9 or newer is required for the documented client. The same key is used by the HTTP API, and client construction can alternatively receive the key directly as an argument.

The package provides synchronous and asynchronous clients. Async calls can be combined with asyncio.gather, and each successful 200 response counts as one billed request. Time-series dataclasses can be converted with dataclasses.asdict for workflows such as loading the values into a pandas DataFrame.

Tools and capabilities

• Retrieve a dated Wikipedia attention series for a topic.
• Compare recent, year-to-date, or custom date windows.
• List live Wikipedia trending pages with pagination fields.
• Read normalized values from Python dataclasses or JavaScript objects.
• Request other supported sources through the same client by passing a different source value where applicable.

Limitations and notes

Wikipedia Trends API MCP server requires an API key and is subject to quota limits. A 401 indicates an invalid or missing key, while 429 indicates that the quota was reached. A 404 means that no series exists for the requested keyword and should not be retried. The client retries 5xx failures before raising an error.

Raw curl responses require the JSON string in the HTTP body field to be parsed a second time; the SDKs perform that decoding. The API returns indexed values, not necessarily absolute page-view volume. Absolute volume may be absent from individual time-series points. Use exact article titles when selecting a series, and use type: Wikipedia Trending for the live Wikipedia feed.