# Google Scholar Search API

> The Google Scholar Search API by everydata.io returns live Google Scholar data (scholarly literature search with authors, venue, year, citations and pdf links) as structured JSON: 1 endpoints under /gsc, one API key, 100 free requests a month.

- URL: https://everydata.io/apis/google-scholar
- Updated: 2026-09-24
- Publisher: everydata.io (https://everydata.io)

Search results from Google Scholar as JSON. Query by keyword with an optional year range and result language and get 10 results per page with title, link, result type (article, PDF, book, citation), snippet, authors, venue and year as Scholar displays them, the number of citations with the cited-by link, the number of versions, related articles and the full-text or PDF link. Google Scholar limits bursts from one address: such requests return 502 and are not charged.

## Key facts

- Status: beta
- Endpoints: 1 (path prefix `/gsc`)
- Quota: A call to the Google Scholar API counts as 1 request against the monthly quota.
- Pricing: 100 free requests a month; paid plans from €30 a month for 5,000 requests (€1.00–€6.00 per 1,000 requests). One quota is shared across all 35 platforms; failed requests (5xx, blocked pages) are not counted.
- Measured average response time (last 24 h): 599 ms, success rate 100% – live figures at https://everydata.io/status
- Authentication: `x-api-key` header

## Endpoints

- `GET /gsc/google-scholar-search` – Search scholarly literature. Search Google Scholar (beta). Params: `query`, `page`?, `yearFrom`?, `yearTo`?, `language`?.

## Example

```bash
curl -s "https://api.everydata.io/gsc/google-scholar-search?query=transformer+attention&yearFrom=2018&yearTo=2024&page=2" \
  -H "x-api-key: YOUR_API_KEY"
```

```json
{
  "responseStatus": "PRODUCT_FOUND_RESPONSE",
  "responseMessage": "Product successfully found!",
  "query": "transformer attention",
  "resultCount": 60500,
  "currentPage": 2,
  "organicResults": [
    {
      "position": 11,
      "title": "Contextual attention network: Transformer meets u-net",
      "link": "https://link.springer.com/chapter/10.1007/978-3-031-21014-3_39",
      "type": "ARTICLE",
      "authors": [
        "R Azad",
        "M Heidari",
        "Y Wu",
        "D Merhof"
      ],
      "year": 2022,
      "source": "Springer",
      "citedByCount": 171,
      "versionsCount": 10,
      "pdfLink": "https://arxiv.org/pdf/2203.01932"
    }
  ]
}
```

## Use cases

- Literature reviews and research monitoring
- Citation tracking
- Finding open-access PDFs
- Academic trend analysis

## FAQ

### What data does the Google Scholar Search API return?

Scholarly literature search with authors, venue, year, citations and PDF links. The API has 1 endpoint under the /gsc path – search articles – and every response is structured JSON that starts with responseStatus and responseMessage.

### What does “beta” mean for the Google Scholar API?

The endpoints and response fields are usable today, but they can still change and the platform is not yet counted in the headline success rate. Blocked or failed calls are not billed.

### How much does the Google Scholar API cost?

A call to the Google Scholar API counts as 1 request against the monthly quota. The Google Scholar API is included in every plan and shares one quota with all other platforms: 100 free requests a month; paid plans from €30 a month for 5,000 requests (€1.00–€6.00 per 1,000 requests). Failed requests (5xx, blocked pages) are not counted.

### Is the Google Scholar data realtime?

Yes. Each request loads the live Google Scholar page when you call the API – there is no database of old snapshots in between.

### Do I need a Google Scholar account or an official Google Scholar API key?

No. You only need an everydata.io API key, sent in the x-api-key header. The free plan works without a credit card.

### What happens when Google Scholar blocks a request?

The API retries the fetch automatically. If the page still cannot be loaded you get a 502 or 503 error, and that request costs no quota. Success rates and response times per platform are measured around the clock on the status page.

### Can ChatGPT, Claude or an AI agent use the Google Scholar API?

Yes. Connect the hosted everydata.io MCP server and the assistant can call every Google Scholar endpoint as a tool, or register the plain HTTPS endpoints as functions in any agent framework. The endpoint list is also in /llms.txt and the OpenAPI spec.

## More

- [All platforms](https://everydata.io/apis) · [Pricing](https://everydata.io/pricing) · [API reference](https://everydata.io/docs) · [Getting started](https://everydata.io/docs/getting-started) · [MCP server](https://everydata.io/docs/mcp) · [Status](https://everydata.io/status)
- Machine-readable: [llms.txt](https://everydata.io/llms.txt), [llms-full.txt](https://everydata.io/llms-full.txt), [OpenAPI](https://api.everydata.io/openapi.json)