✂️ PDF Page Splitter

Split any PDF into individual pages — then chain into 100+ Apify actors for extraction, OCR, or analysis.

What does PDF Page Splitter do?

PDF Page Splitter takes a multi-page PDF and splits it into individual single-page PDFs. You can extract all pages, pick specific pages (e.g., 1,3,5), or select a range (e.g., 1-5).

This is the essential first step in any document processing pipeline.

Key features

• ✅ Flexible page selection — all, 1, 1,3,5, 1-5, 1,3-5,7
• ✅ Two output modes — Dataset (base64 JSON for chaining) or Key-Value Store (binary files for download)
• ✅ Upload or URL — Upload a PDF directly or provide a URL
• ✅ Memory efficient — Processes pages one-by-one with periodic cleanup
• ✅ Lightweight — Runs on just 256 MB of memory
• ✅ Handles massive PDFs — Up to 50,000 pages / 500 MB
• ✅ Flat pricing — One flat fee per run, no matter how many pages

---

Use cases

Use Case	Pages Setting	Output Format
Split batched invoices for processing	all	Dataset (base64)
Extract a specific contract page for review	3	Key-Value Store
Pull first 5 pages from a report	1-5	Dataset (base64)
Cherry-pick pages from a large document	1,5,10-15,20	Key-Value Store

Who is this for?

• n8n / Make.com automation builders who process documents in workflows
• Finance teams splitting batched invoices and payment advices
• Legal teams extracting specific pages from large contracts
• AI/RAG pipeline developers who need individual pages for processing

---

Input

Field	Type	Default	Description
pdfFile	File Upload	—	Upload a PDF file directly
pdfUrl	String	—	URL to a PDF file (for chaining from other actors)
pages	String	all	Page selection: all, 1, 1,3,5, 1-5, 1,3-5,7
outputFormat	Enum	dataset_base64	dataset_base64 (JSON with base64) or key_value_store (binary files)

Note: Provide either pdfFile or pdfUrl. If both are provided, pdfFile takes priority.

Page selection examples

all          → Extract every page
1            → Just the first page
1,3,5        → Pages 1, 3, and 5
1-5          → Pages 1 through 5
1,3-5,7      → Pages 1, 3, 4, 5, and 7

---

Output

Dataset output (dataset_base64)

Each page becomes one row in the dataset:

{
    "page_number": 1,
    "filename": "page_1_invoice.pdf",
    "size_bytes": 12345,
    "content_base64": "JVBERi0xLjQK...",
    "original_filename": "invoice.pdf",
    "total_pages": 10
}

Key-Value Store output (key_value_store)

Each page is saved as a binary PDF file in the default Key-Value Store:

• page_1_invoice.pdf
• page_2_invoice.pdf
• page_3_invoice.pdf
• etc.

---

Pricing

This Actor uses flat per-run pricing:

Event	Price
run-completed (per run, any number of pages)	$0.005

Example costs:

• Split a 10-page PDF → $0.005
• Extract 3 specific pages → $0.005
• Split a 10,000-page document → $0.005

Same price whether you split 1 page or 10,000. Platform compute costs vary by memory and run time.

---

Chaining with other actors

PDF Page Splitter is designed as the gateway to document processing pipelines:

┌─→ Resume Text Extractor
                    │
PDF Page Splitter ──┼─→ Indian Payment Advice Parser
                    │
                    ├─→ Document Table Extractor
                    │
                    └─→ PDF to PNG Converter

Input sources

• Any file download actor
• Gmail attachment scraper
• Website crawler (PDF links)
• Manual upload

Output destinations

• Resume Text Extractor — Extract structured text from resume PDFs
• Indian Payment Advice Parser — Parse payment details from bank advices
• Document Table Extractor — Extract tables from document pages
• PDF to PNG Converter — Convert pages to images for OCR or vision AI

---

Integration with n8n / Make.com

n8n HTTP Request Node

1. Set Method to POST
2. Use the Apify API endpoint to run this actor
3. Pass your PDF as binary data or provide a URL
4. Use dataset_base64 output for easy downstream processing

Make.com

1. Use the Apify module to run this actor
2. Pass your PDF URL in the pdfUrl input field
3. Iterate over dataset results for downstream processing

---

Limits

Resource	Limit
File size	500 MB
Page count	50,000 pages
Default memory	256 MB
Max memory	4096 MB

Memory recommendations

PDF Size	Pages	Recommended Memory
≤ 10 MB	≤ 100	256 MB
≤ 50 MB	≤ 500	512 MB
≤ 200 MB	≤ 2,000	1024 MB
≤ 500 MB	≤ 5,000	2048 MB
> 500 MB	> 5,000	4096 MB

Tip: For PDFs with 1,000+ pages, use key_value_store output format instead of dataset_base64 to avoid base64 encoding overhead.

---

Changelog

v0.2 (Large PDF Support)

• Flat per-run pricing ($0.005 per run, unlimited pages)
• Raised limits: 50,000 pages, 500 MB file size, 4 GB memory
• Streaming download for large files
• Batch dataset pushes (50 rows per API call)
• Aggressive memory cleanup with gc for 10K+ page PDFs
• Progress logging every 100 pages
• Memory recommendations in logs

v0.1 (Initial Release)

• PDF splitting with page selection (all, ranges, specific pages)
• Two output modes: Dataset (base64) and Key-Value Store (binary)
• File upload and URL input support
• Memory-efficient processing with periodic cleanup