Handwriting to Text OCR: Notes, Letters, Forms & PDFs

Turn photos, scans and PDFs of handwriting into text you can copy, search and edit, page by page. Paste links to your files (JPG, PNG, HEIC iPhone photos, PDF…, or Google Drive and Dropbox file links) and get, for every page:

• ✍️ Exact transcription with line breaks: the text is copied exactly as written, line by line, in the original spelling. Nothing is corrected, summarised or translated.
• 🌍 Any language: including non-Latin scripts such as Cyrillic, Greek, Arabic, Chinese, Japanese and Hindi, with the detected language as a BCP-47 code.
• ❌ Crossed-out words dropped, unreadable words marked: words that were struck through are left out, and words that can't be read are marked [illegible] and counted.
• 📄 PDFs and multi-page TIFFs page by page: one record per page, with the page number and page count.
• 📱 HEIC iPhone photos and auto-rotation: phone photos are turned the right way up from their orientation data before reading.
• 🔢 Word and line counts per page, counted by code from the text.

No API key, no account to connect, no setup. It works for notes, notebooks, letters, forms, whiteboards and old documents.

What can you do with it?

• Digitise meeting notes and notebooks: make handwritten notes searchable and paste them into your notes app or task list.
• Letters and family archives: transcribe old letters, postcards and diaries so the whole family can read and keep them.
• Handwritten forms and surveys: get the answers into text for spreadsheets, CRMs and databases.
• Whiteboard photos: keep what was written in a workshop or class as text instead of a blurry photo.
• Historical and genealogy documents: a first transcription of registers, minutes and records to proofread and search.
• Exam answer sheets: prepare handwritten answers as text for teachers to read and grade faster.
• AI assistants and RAG: feed handwritten notes into your AI assistant, search index or knowledge base.

What files does it accept?

Input	Example	Supported
Direct image file link	https://example.com/scans/letter.jpg	✅ JPG, PNG, WebP, HEIC/HEIF (iPhone), TIFF incl. multi-page, GIF, BMP
Direct PDF link	https://example.com/scans/notebook.pdf	✅ scanned or photographed pages, rendered at 200 dpi
Cloud storage file URL	a public file URL from S3 or another bucket	✅ if it opens without logging in
Google Drive file	https://drive.google.com/file/d/…/view	✅ shared as "Anyone with the link"
Dropbox file	https://www.dropbox.com/scl/fi/…?dl=0	✅
Folders, or pages that need a login		❌ share each file's own link

By default up to 20 pages are transcribed per file; raise Max pages per file (up to 200) for long PDFs, or lower it to cap the cost of long ones.

How much does it cost?

Pay per page. No monthly fee.

Event	Price (FREE/BRONZE)	Lower on paid plans
Page transcribed (text, language, handwritten flag, counts)	$0.005 / page	down to $0.004
Actor start	$0.00005 / run	—

100 pages cost $0.50; a 20-page PDF costs $0.10. You pay only for pages actually transcribed. Nothing is charged for a page or file that fails.

Cap your spend: set a maximum cost per run. When the budget runs out, no more pages are transcribed.

Free plan: the AI runs on paid infrastructure, so free-plan runs include a 3-page trial per run; upgrade to transcribe more pages.

How to convert handwriting to text

1. Click Try for free. The example input transcribes a sample handwritten note.
2. Paste your own links into Photo, scan or PDF links, one per line.
3. Optional: set Max pages per file for long PDFs.
4. Run, then open the Pages or Text view, or download JSON, CSV or Excel.

Input example

{
  "fileUrls": [
    "https://example.com/scans/letter.jpg",
    "https://example.com/scans/notebook.pdf"
  ],
  "maxPagesPerFile": 50
}

Output

One record per page. A page of meeting notes:

{
 "url": "https://example.com/scans/notes.pdf",
 "source": "direct",
 "page": 1,
 "pageCount": 2,
 "text": "Team sync - Tuesday 14 Oct\n- Launch moved to Nov 3 (waiting on legal)\n- Maria owns the pricing page copy\n- Budget: $12,400 left for Q4 ads\n- Ask Tom about the Berlin office lease\nNext call: Friday 9:30 am",
 "language": "en",
 "handwritten": true,
 "illegibleWords": 0,
 "lineCount": 6,
 "wordCount": 42,
 "processedAt": "2026-10-15T08:00:00+00:00"
}

text keeps the line breaks of the page. handwritten is true when most of the text on the page is handwritten. illegibleWords, lineCount and wordCount are computed by code from the text, not guessed by the AI.

How accurate is it?

Measured with Google's Gemini 3.1 Flash-Lite model on:

• 4 synthetic handwritten notes (made with handwriting fonts, with jitter, rotation and blur): every character right on all 4.
• 12 real phone photos of handwritten notes from a public CC0 dataset: 11 word-for-word exact; the 12th differed only in capitalisation ("ocr" vs "OCR") after crossed-out words were left out.
• A 1916 Danish handwritten minutes page (old cursive, Aarhus City Archives, CC BY 4.0): about 84% of characters right.

Tips: use clear, well-lit photos, and one page per photo. For old scripts, expect to proofread.

How fast is it?

Pages are transcribed 4 at a time. Measured on Apify at the default 1 GB: a single photo in about 8 seconds including start-up, and a 20-page PDF in 14 seconds.

Use it with AI agents (MCP)

https://mcp.apify.com?tools=bossless/handwriting-to-text

Ask your assistant "what does this handwritten note say?", and it runs this actor and reads back the text.

Use the API from code

curl -X POST "https://api.apify.com/v2/acts/bossless~handwriting-to-text/run-sync-get-dataset-items?token=YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"fileUrls": ["https://example.com/scans/letter.jpg"]}'

from apify_client import ApifyClient

client = ApifyClient("YOUR_TOKEN")
run = client.actor("bossless/handwriting-to-text").call(run_input={
    "fileUrls": ["https://example.com/scans/notebook.pdf"],
    "maxPagesPerFile": 50,
})
for item in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(f"--- page {item['page']} of {item['pageCount']} ({item['wordCount']} words)")
    print(item["text"])

FAQ

Do I need a Google, OpenAI or other API key? No. Everything is included in the price.

Which languages work? Most languages and scripts, including Cyrillic, Greek, Arabic, Chinese, Japanese and Hindi. The text stays in the language it was written in, and language tells you which one was detected. Quality varies with the handwriting.

Does it read printed text too? Yes. Printed text on the page is transcribed as well; handwritten tells you whether handwriting or print dominates the page.

What about tables and forms? They come out as lines of text in reading order. Cell structure is not kept.

What happens to crossed-out words? They are left out, so the text reads the way the writer meant it.

What if a word can't be read? It's marked [illegible] in the text and counted in illegibleWords.

What happens with a private or broken link? It's skipped with a clear message and not charged.

Is my data stored or used for training? Files are downloaded only for the run and deleted when it ends. Results are saved in your own Apify dataset. The AI model runs on Google Cloud (Vertex AI). Google's terms say it won't use this data to train or fine-tune its models; it may cache requests for up to 24 hours and log them for abuse monitoring.

How are pages counted? One photo or image is one page; a PDF or multi-page TIFF counts each page transcribed, up to Max pages per file.

Is this an official product of any provider? No. This is an independent tool and is not affiliated with or endorsed by Google, Apple, Microsoft or any OCR provider.

Related tools

• Speaker Diarization & Transcription: who spoke when in any audio or video, with a transcript with speaker names, SRT/VTT subtitles and RTTM.
• Call Recording Transcription & AI Call Analysis: speaker-labelled call transcripts with an AI summary, sentiment, action items and a QA scorecard.