Every PDF job, one studio.
Documents is PDF Studio in Atlas: 39 tools to organize, convert, edit, secure, sign, compress, and read PDFs with AI. Most tools run client-side in your browser. eSign and OCR run inside your Atlas workspace, never through a third-party uploader.
Eight groups, 39 tools.
PDF Studio works on PDFs and converts to and from Word, Excel, PowerPoint, images, and HTML. OCR makes scanned PDFs searchable, and the smart tools read what is on the page.
Merge, split, reorder pages, extract pages, insert pages or blank pages, delete pages, rotate, crop, and add page numbers.
PDF to Word, Excel, PowerPoint, JPG, PNG, ZPL, or PDF/A. Word, Excel, PowerPoint, JPG, HTML, ZPL, or a scan back to PDF.
Edit text, watermark, fill and edit forms, compare two PDFs, and annotate. Add Bates numbers, redact for good, protect with a password, unlock files you own, and repair a damaged PDF.
Sign a PDF yourself or request signatures from other people, and compress large files while keeping the best possible quality.
OCR turns scanned PDFs into searchable, selectable text. Summarize writes a plain-language summary of a PDF. Translate converts a PDF into another language and keeps the original layout.
Three steps, in order.
A typical job has three steps: get the text, clean up what should not be there, and read it with AI. You choose which steps a file needs.
OCR and text
OCR adds a searchable text layer to scanned pages, in up to three languages per run. Table extraction pulls tables out as data.
Find what to remove
PII discovery flags personal data for review. Redaction then removes the content from the file permanently, not just from view.
Ask, summarize, translate
AI Q&A, summary, extract, and translate run on the documents you choose. The summary refuses image-only scans rather than guess, so run OCR first.
Nine more, all in production.
Beyond the 39 tools, PDF Studio and the Atlas API also handle these.
PDF jobs
Run work as a job, follow live progress, cancel it, then download or archive the result.
Encrypted throughout
TLS 1.3 in transit and encryption at rest from our managed host. Your content is never used to train a model.
Document versions
Commit a new version of a document, download any version, and restore an older one.
OCR in 10 languages
English, Spanish, Hindi, French, German, Japanese, Chinese, Arabic, Portuguese, and Hebrew.
Annotation threads
Annotations carry comment threads that teammates can resolve or reopen.
Bates numbering
Stamp sequential Bates numbers on every page, with your prefix and start number.
PKCS#7 signatures
Apply a PKCS#7 digital signature to a PDF, or verify the signatures already on one.
PDF/A
Preflight a file, then convert it to PDF/A for long-term archiving.
Form fill
Fill AcroForm fields, import or export FDF, then flatten the result.
Documents flows back in.
A document is rarely a destination. The output usually needs to land as a task, a contract, a row in CRM, or a doc.
Five places a PDF connects
Contracts and eSign handle signing with an audit trail. PDF Studio documents have their own search index. Follow-up work goes into Tasks. Automations can generate a PDF with AI as an action, and the CRM shows quotes, proposals, and contracts inline.
Programmatic documents.
OCR, ask, summarize, convert, or stamp over the REST API. Files go in as base64, up to 50 MiB each.
# Encode the file once. PDF=$(base64 < agreement.pdf | tr -d '\n') # Add a text layer to a scanned PDF (up to three OCR languages). curl -X POST https://api-atlas.wrxstack.com/v1/pdf-ocr/extract \ -H "Authorization: Bearer atlas_pat_..." \ -H "Content-Type: application/json" \ -d "{ \"pdfBase64\": \"$PDF\", \"languages\": [\"eng\", \"deu\"] }" # Ask a question grounded in the document (the answer streams back). curl -X POST https://api-atlas.wrxstack.com/v1/pdf-intelligence/ask \ -H "Authorization: Bearer atlas_pat_..." \ -H "Content-Type: application/json" \ -d "{ \"pdfBase64\": \"$PDF\", \"question\": \"Where can we terminate for convenience?\" }" # Stamp Bates numbers for a legal production set. curl -X POST https://api-atlas.wrxstack.com/v1/pdf-legal/bates \ -H "Authorization: Bearer atlas_pat_..." \ -H "Content-Type: application/json" \ -d "{ \"pdfBase64\": \"$PDF\", \"prefix\": \"PROD-\", \"start\": 1 }"
Documents vs. Google Document AI.
Google Document AI is a set of cloud services for building document pipelines. Atlas PDF Studio is a ready-to-use set of PDF tools inside your workspace.
wrxstack Documents vs. Google Document AI
| Capability | Google Document AI | wrxstack Documents |
|---|---|---|
| End-to-end product | Cloud APIs you assemble | Browser tools plus a REST API |
| Everyday PDF tools | Not its focus | Merge, split, convert, compress, and more |
| Q&A on a document | Build it with other services | AI Q&A built in |
| Compare two versions | Not its focus | Visual diff of two PDFs |
| eSign | Not provided | eSign with a completion certificate |
| Custom document parsers | Custom processors at cloud scale | AI extract for structured fields |
Common questions.
Six things legal, claims, and procurement teams ask before adopting Documents.
Can the AI get a summary wrong?
Yes, AI output can contain mistakes. The AI summary is grounded in the document you give it, and it refuses an image-only scan rather than guess, so run OCR first. Review AI output before you rely on it.
How is our document data secured?
Most PDF Studio tools run client-side in your browser. eSign and OCR run inside your Atlas workspace, never through a third-party uploader. Data travels over TLS 1.3 and is encrypted at rest by our managed hosting provider. Your content is not used to train any model.
What about scanned PDFs?
Run OCR to add a searchable text layer, in up to three of ten supported languages per run. AI summary, Q&A, and extract read the text layer, so OCR comes first for image-only pages.
How do larger jobs run?
Larger work runs as a PDF job in your workspace. You can follow live progress, cancel a job, and download or archive the result when it finishes. API calls accept files up to 50 MiB each.
Which file types can I convert?
PDF to Word, Excel, images (PNG or JPG), and HTML, and Word, Excel, images, and HTML to PDF. Through the API, a URL can also be fetched as a PDF source.
Can extraction match our forms?
AI extract pulls structured fields, and table extraction returns a PDF's tables. For fillable PDFs, read the AcroForm field values directly or export them as FDF, with no AI involved.
Pairs well with.
PDF Studio pairs with Contracts for signing, Search for finding files, and Tasks for the follow-up.
Who this is not for.
Documents organizes, converts, edits, secures, signs, and reads PDFs. If you are a law firm or records team that needs a full document-management system with matter-centric filing, retention schedules, and legal hold, a dedicated DMS will fit better. Documents processes and understands files, it is not a system of record for regulated retention.
Open a PDF right now.
Free to start, no credit card. Pricing is set inside the product, and the AI suite uses Atlas private AI, which is on paid plans.