# Google Scholar

`POST https://api.serpkite.com/v1/scholar` · Credits: 1 per page

Papers with authors, publication, year, citation count and PDF links.

Google Scholar papers with authors and publication, year, citation count and a direct PDF link when one exists.

## Request body

| Parameter | Type | Default | Description |
| --- | --- | --- | --- |
| `q` **required** | string |  | The search query. Required. Up to 2,048 characters. |
| `language` | string | `en` | Interface language, as a language code (en, de, fr, pt-BR…). |
| `num` | integer | `10` | Results per call. 10 per page; 100 fetches the top 100 as a depth bundle for 7 credits instead of 10. One of: `10`, `20`, `30`, `50`, `100`. |
| `page` | integer | `1` | Results page, 1–10. Each page is billed separately. |
| `format` | string | `json` | Response format. markdown is LLM-ready prose; compact is JSON with only the fields agents need. One of: `json`, `compact`, `markdown`. |
| `fields` | string |  | Comma-separated projection, e.g. results.title,results.link,knowledge_graph. Cuts tokens. |
| `max_age` | integer |  | Accept a cached result up to this many seconds old. Cache hits cost 50% of the credits. |

## Example request

cURL:

```bash
curl https://api.serpkite.com/v1/scholar \
  -H "Authorization: Bearer $SERPKITE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"q":"retrieval augmented generation"}'
```

TypeScript:

```ts
import { SerpKite } from "serpkite";

const sk = new SerpKite(); // reads SERPKITE_API_KEY
const res = await sk.scholar({ q: "retrieval augmented generation" });
console.log(res.results[0].title, res.meta.credits_used);
```

Python:

```python
from serpkite import SerpKite

sk = SerpKite()  # reads SERPKITE_API_KEY
res = sk.scholar("retrieval augmented generation")
print(res.results[0].title, res.meta.credits_used)
```

## Response fields

| Field | Type | Description |
| --- | --- | --- |
| `request` | object | The normalised request, with defaults filled in: `endpoint`, `engine`, `q`, `country`, `language`, `location`, `num`, `page`, `device`, `autocorrect`… |
| `results[]` | array | `position`, `title`, `link`, `domain`, `publication_info`, `snippet`, `year`, `cited_by`, `pdf_url`, `id`. |
| `meta` | object | `request_id`, `credits_used`, `cached`, `cached_at`, `engine` (the provider that answered), `route` (provider attempts, see [Search providers](https://serpkite.com/docs/providers)), `latency_ms`, `parse_quality` (`ok`, `partial`, `empty`), `resolved_urls`. |

## Example response

```json
{
  "request": {
    "endpoint": "scholar",
    "engine": "google",
    "q": "retrieval augmented generation",
    "language": "en"
  },
  "results": [
    {
      "position": 1,
      "title": "Retrieval-augmented generation for knowledge-intensive NLP tasks",
      "link": "https://papers.example.org/rag",
      "domain": "papers.example.org",
      "publication_info": "P Lewis, E Perez, A Piktus… - Advances in Neural Information Processing Systems, 2020",
      "snippet": "Large pre-trained language models have been shown to store factual knowledge…",
      "year": 2020,
      "cited_by": 9000,
      "pdf_url": "https://papers.example.org/rag.pdf",
      "id": "abcDEF123"
    }
  ],
  "meta": {
    "request_id": "req_01J8ZK4M6Q2V7",
    "credits_used": 1,
    "cached": false,
    "engine": "google",
    "latency_ms": 1034,
    "parse_quality": "ok",
    "resolved_urls": true
  }
}
```

## Errors

| Status | Code | Meaning |
| --- | --- | --- |
| 400 | `invalid_request` | A parameter is missing or invalid. |
| 401 | `unauthorized` | The API key is missing, invalid or revoked. |
| 402 | `insufficient_credits` | Your balance is too low. Buy a pack or wait for the monthly free grant. |
| 429 | `rate_limited` | Too many requests per second for your plan. Retry after the Retry-After header. |
| 503 | `upstream_error` | Google could not be fetched or parsed. Not billed; retry after Retry-After. |

All errors: https://serpkite.com/docs/errors

## Related

- [RAG pipeline](https://serpkite.com/docs/guides/rag-pipeline)