# Cleanup JSON (https://crawler-docs.lumen-labs.ai/docs/v2/cleanup)

> For the complete Lumen Crawler documentation index, see [llms.txt](https://crawler-docs.lumen-labs.ai/docs/llms.txt). A markdown version of any docs page is available by appending `.md` to its URL (e.g. https://crawler-docs.lumen-labs.ai/docs/v2/cleanup.md).

Per-page and site-level typed JSON after the crawl. Default on. Second wait after status.

Cleanup runs after fetches when `cleanup` is true (the default). It classifies main content versus chrome and builds a site rollup (areas, headings).

Crawl `status` can be `completed` while `cleanup.status` is still `pending` or `running`.

## Per-page cleaned

```bash
curl "https://crawler.lumen-labs.ai/v1/crawls/$ID/pages?include=cleaned" \
  -H "Authorization: Bearer $LUMEN_CRAWLER_KEY"
```

Combine with markdown: `include=markdown,cleaned`.

## Site rollup

```bash
curl https://crawler.lumen-labs.ai/v1/crawls/$ID/cleaned \
  -H "Authorization: Bearer $LUMEN_CRAWLER_KEY"
```

| HTTP | `type` | Meaning |
| --- | --- | --- |
| 200 | — | Rollup ready |
| 409 | `cleanup_not_ready` | Cleanup still queued or running |
| 404 | `cleanup_unavailable` | Cleanup skipped, failed, or missing |

`cleanup` on the crawl object has `status` (`skipped`, `pending`, `running`, `completed`, `failed`), `model`, `pages_cleaned`, and `error`.

Set `"cleanup": false` on create to skip. Then `/cleaned` is 404.

<AgentNote>

After crawl `status` is terminal, keep polling GET `/cleaned` until 200 or 404. Treat 409 as wait, not failure.

</AgentNote>