PDF Text Extractor
Pricing
from $1.50 / 1,000 document with extracted texts
PDF Text Extractor
Extract text from PDF URLs as JSON: full text, per-page text, metadata and optional tables, one row per document. Flags scanned image-only PDFs. From $2.00 per 1,000 extracted documents, platform usage included.
Pricing
from $1.50 / 1,000 document with extracted texts
Rating
0.0
(0)
Developer
subimpact
Maintained by CommunityActor stats
0
Bookmarked
1
Total users
1
Monthly active users
14 hours ago
Last modified
Categories
Features
Share
Turn a list of PDF URLs into structured JSON: the full document text, per-page text, PDF metadata and, optionally, tables — one row per document, delivered to the run's dataset.
No browser, no API key, no login. Plain HTTP download plus a PDF parser, so a run over dozens of documents finishes in seconds.
What you get
For each PDF URL:
text— the whole document as plain text with normalised whitespace (capped at 1,000,000 characters, flagged withtruncated).pageTexts— one text string per extracted page, so you can cite "page 4" without re-parsing.tables— optional table extraction as arrays of rows (includeTables).title,author,subject,creator,producer,createdAt,modifiedAt— the PDF's own metadata, dates normalised to ISO 8601 UTC.pageCount/pagesExtracted— real page count, and how many pages were read (capped bymaxPages).isScanned—truewhen the document has almost no text layer (an image-only scan). Those rows are delivered but not charged, because v1 does not run OCR.status—ok(text found),scanned,emptyorerror, plus anerrormessage when a download or parse failed.
Typical uses
- Search and RAG pipelines that need clean text from reports, filings, contracts, forms and papers.
- Document intake: detect digital vs scanned PDFs before sending a file to an OCR service.
- Pull a specific table out of a recurring PDF report (invoices, statements, regulatory tables).
- Archive text alongside the source URL, page count and metadata.
Input
{"urls": [{ "url": "https://www.irs.gov/pub/irs-pdf/fw9.pdf" },{ "url": "https://nvlpubs.nist.gov/nistpubs/Legacy/SP/nistspecialpublication800-145.pdf" },"https://arxiv.org/pdf/1706.03762"],"maxPages": 200,"includeTables": false,"includePageTexts": true,"maxDocuments": 100,"maxFileSizeMb": 50}
| Field | Default | Meaning |
|---|---|---|
urls | two small public PDFs | PDF links. A plain string or {"url": "..."} both work; redirects are followed. |
maxPages | 200 | Pages read per document. pageCount still reports the real total. |
includeTables | false | Add tables ([{"page": 1, "rows": [["cell", ...], ...]}]). Slower than text-only. |
includePageTexts | true | Add pageTexts (one string per extracted page). Turn off for a smaller dataset. |
maxDocuments | 100 | Hard cap on delivered rows per run, including scanned/empty/error rows. Maximum 5,000. |
maxFileSizeMb | 50 | Larger files are skipped as an error row; the download is aborted at the limit. |
proxyConfiguration | off | Optional Apify Proxy, only needed if a publisher blocks the run's IP. |
Output
A real row from a live run over the IRS Form W-9 (text shortened here; the dataset row carries the full 37,694 characters):
{"url": "https://www.irs.gov/pub/irs-pdf/fw9.pdf","finalUrl": "https://www.irs.gov/pub/irs-pdf/fw9.pdf","fileName": "fw9.pdf","sizeBytes": 140815,"pageCount": 6,"pagesExtracted": 6,"title": "Form W-9 (Rev. March 2024)","author": "SE:W:CAR:MP","subject": "Request for Taxpayer Identification Number and Certification","creator": "Designer 6.5","producer": "Designer 6.5","createdAt": "2024-03-06T13:18:13Z","modifiedAt": "2024-03-06T13:18:13Z","text": "Form W-9\n(Rev. March 2024)\nRequest for Taxpayer\nIdentification Number and Certification\nDepartment of the Treasury\nInternal Revenue Service ...","truncated": false,"pageTexts": ["Form W-9\n(Rev. March 2024)\nRequest for Taxpayer\nIdentification Number and Certification ...","Form W-9 (Rev. 3-2024) Page 2\nmust obtain your correct taxpayer identification number ..."],"tables": null,"isScanned": false,"language": "en","scrapedAt": "2026-11-15T09:08:29Z","status": "ok","error": null}
Every field is declared in the dataset schema, so the dataset table, CSV and Excel exports show all of them.
Pricing
$2.00 per 1,000 documents — one charge per document delivered with extracted text (status: ok). Platform usage is included.
Scanned, empty and error rows are delivered with their status and are free: you only pay when a document actually yields text. At the default maxDocuments: 100 a full run costs at most $0.20.
Other PDF text-extraction APIs on the Apify Store list between $3.45 and $5.00 per 1,000 documents (checked November 2026); this one is $2.00.
Notes and limits
- No OCR in v1. Image-only PDFs are detected (
isScanned: true,status: scannedorempty) and not charged. Send those to an OCR actor if you need their text. - Password-protected PDFs cannot be opened and come back as
status: errorwith a clear message. - Text quality depends on the PDF. Documents built from text extract cleanly; unusual font encodings can lose characters.
fontToolsis installed to read CFF/Type1 fonts properly. - Whitespace is normalised: runs of spaces/tabs collapse and blank lines are dropped, so layout columns are not preserved. Use
tablesfor tabular data andpageTextsfor page boundaries. textis capped at 1,000,000 characters;truncated: trueandpageTexts/pagesExtractedtell you what was cut.- Downloads are sequential and streamed, one document at a time, with a size cap — memory use stays around 100 MB in a typical text-only run.
Run monitoring and privacy
If a run fails or times out, this Actor tells its developer the run ID and status, so we can fix problems fast. It never sends your input, your results or your account data.