Document OCR to JSON: images and PDFs to text, tables
Pricing
$2.00 / 1,000 one page or image of extracted texts
Document OCR to JSON: images and PDFs to text, tables
Turn image and PDF URLs into clean text, key-value pairs and table rows as JSON, one row per page.
Pricing
$2.00 / 1,000 one page or image of extracted texts
Rating
0.0
(0)
Developer
Maybe Quark
Maintained by CommunityActor stats
0
Bookmarked
2
Total users
1
Monthly active users
a day ago
Last modified
Categories
Share
Document OCR to JSON
Document OCR to JSON turns image and PDF URLs into structured JSON: the page text, detected key-value pairs (such as "Invoice No: 1042"), table-like blocks as rows of cells, a word count and an OCR confidence score. It is built for AI agents, automation builders and developers who need a cheap, predictable text-extraction step in a pipeline. One output row is one page or one image.
Who it is for
- AI agents and MCP clients that need to read a receipt, form or scan.
- Developers feeding documents into RAG or search indexes.
- Ops teams extracting fields from invoices and statements.
What each row has
| Field | Meaning |
|---|---|
| url | Source document URL |
| page | Page number (1 for images) |
| source_type | image or pdf |
| status | ok, empty, no_text_layer or error |
| text | Extracted text |
| word_count | Words found |
| confidence | OCR confidence 0-100 for images, null for PDFs |
| key_values | Object of Label: value pairs |
| tables | Array of tables, each an array of rows of cells |
| error | Reason when status is not ok |
Sample row:
| url | page | source_type | status | word_count | confidence |
|---|---|---|---|---|---|
| (image URL) | 1 | image | ok | 120 | 91 |
How to use
Paste public image or PDF URLs into Document URLs, pick the OCR language (default eng), and run. With the API:
curl -X POST "https://api.apify.com/v2/acts/~document-ocr-to-json/run-sync-get-dataset-items?token=YOUR_TOKEN" -H "content-type: application/json" -d '{"documentUrls":["https://example.com/receipt.png"],"language":"eng"}'
It can also be called as a tool through the Apify MCP server, so an agent can pass document URLs and read back the JSON rows.
Use cases
- Read scanned receipts and invoices into fields.
- Convert text PDFs into line-preserved text for RAG.
- Pull tabular blocks out of statements and reports.
Limits and FAQ
Are scanned PDFs supported? Not yet. PDFs need a text layer. A scanned page returns status no_text_layer; submit it as an image URL instead.
Are table results perfect? No. Tables are detected from column gaps and work best on clean layouts.
Which languages? Any Tesseract language code for images.
Handwriting? Poor results; printed text only.
Pricing
Pay per result: $0.002 per row (one page or image). Failed downloads also produce a row, so check the status field.
Data source
Only the documents you supply by URL. Text is recognised locally with Tesseract (tesseract.js) and the pdf.js text layer. No logins, nothing stored beyond your dataset.
Related compliance tools
Other Maybe Quark actors: contractor-license-verifier, fmcsa-carrier-safety-lookup, secretary-of-state-business-search, healthcare-license-verifier, npi-registry-lookup, exclusion-debarment-screener, fda-recalls.
