Help center · AI Sortify

Why split a PDF?

A scanner may save invoices, letters, and contracts together. Splitting gives each document its own file so you can find, share, and organize it separately. A multi-page document should usually stay together.

1. Add your source PDF

Web: use Add file or Upload PDF to choose a PDF from your computer.
Desktop or Split Companion: choose a watched folder for incoming PDFs and an output folder for generated copies. A browser alone does not continuously watch your computer’s folders.
Before starting: check the displayed file-size limit, PDF Split allowance, and workspace access. Keep the original available until you have checked the results.

2. Start analysis

Manual processing: start the detected file with Split PDF, then open Review split when analysis finishes.
Uploading: the source is being transferred for processing.
Analyzing: AI checks the pages and proposes document boundaries. These are suggestions to review, not a guarantee that every boundary is correct.
Progress: keep the web page open during active work. View progress opens the current job without starting another one.

3. Review names and page groups

Preview: compare the source pages with each proposed document. Keep continuation pages, attachments, and signatures with the document they belong to.
Edit a group: change its document name and assigned pages using the editing controls. Use a name that helps you recognize the result.
Add page group: create a document from unassigned pages when the detected groups miss a separate document.
Before confirmation: every source page must be assigned, and no page may appear in more than one group. Resolve missing-page or duplicate-page messages first.

4. Confirm and create PDFs

Confirm & create PDFs accepts the reviewed groups and creates the child PDFs. Wait until creation finishes before using the result actions.

Web results: use Download ZIP when available to save the generated PDFs together.
Local results: generated copies are saved to the selected output folder.
Original: creating child PDFs does not mean the original was replaced. Local deletion is controlled separately by the cleanup settings.

5. Send results for organization

Creating split PDFs and classifying documents are separate steps. Use the available upload or confirmation action to pass results to the workspace processing flow.

Split names: help identify each output during review.
Classification: determines the document type and extracts information when the outputs enter document processing.
Final organization: follows your workspace folder structure, folder behavior, custom fields, and file-naming settings. A split result is not proof that cloud filing has completed.

Choose how much to automate

Auto process split files: starts upload and analysis for detected PDFs. Multiple files wait in a queue and run one at a time.
Auto-confirm split results: accepts detected groups and creates PDFs without your review. Leave this off when you want to check boundaries first.
Auto-upload after split: uploads generated PDFs after automatic confirmation. This setting requires Auto-confirm.
Manual review: leave Auto-confirm off while learning how your scans are grouped. Enable more automation only when it matches your workflow.

Fast, Balanced, or Accurate?

Choose the split mode in Split settings. Review the proposed groups in any mode, especially when documents look similar.

Fast: favors shorter analysis time.
Balanced: recommended for most mixed PDFs.
Accurate: spends more time checking difficult document boundaries.

Watched folders and Split Companion

A connected computer runs the local watched-folder workflow. The watched folder contains incoming PDFs; the output folder stores generated copies.

Connected: the computer is available for its local workflow.
Offline or paused: check that the app is running, the computer is connected, and watching is enabled before expecting new files to process.
No computer connected: manual web upload remains a separate way to submit a PDF.
Folder access: if files do not appear, check the selected folder and the app’s access to it. Admin diagnostics can show watcher and queue details.

Upload from your phone and resume jobs

Upload from phone: scan the QR code, sign in with the same account, and upload your PDF. Follow the size limit shown in the upload screen.
Expired code: choose New QR code to continue.
Synced jobs: open a ready job from your devices to review its groups or save its PDFs on the current device. If the source is unavailable, follow the prompt to provide the original.

Understand the file statuses

Detected: the file has been found and is waiting to start.
Queued, uploading, analyzing: the job is waiting or actively preparing groups.
Split ready: open the proposed groups and review them.
Split confirmed: groups have been accepted; check the available output and upload actions.
Done: the split workflow has completed. Check Documents separately for the state of any subsequent classification.
Failed or canceled: read the job message and use Try again when available. Check existing results before submitting the same source again.

Duplicates, limits, and unavailable originals

Duplicate warning: Ask before adding duplicate files warns about previously detected content or the same name and size.
Ignore duplicate detection: allows matching PDFs to process and overrides the duplicate warning setting. Enable it only when repeated processing is intended.
Plan allowance: PDF Split has its own usage limit. Check the current usage and any access or size error before retrying; classification is a separate workflow.
Restore a saved job: if the original is unavailable, choose the original PDF again when prompted so the app can create the output files. Saved group information alone does not contain all source pages.

History and cleanup are different

Local cleanup controls are available in the local workflow. Check which copies you want to keep before enabling deletion.

Completed split history: controls how many completed records stay visible. It does not delete the original or generated PDFs.
Clean generated files: removes tracked generated PDFs from the selected local output folder.
Reset Split workspace: removes tracked generated PDFs and clears local Split history and the duplicate cache. Original watched-folder files and cloud files are kept.
Auto-delete original: when enabled, deletes the local source after all generated files upload successfully. It is off by default.
Auto-delete generated: when enabled, deletes user-facing local output copies after successful upload. It is off by default.

A simple first run

Choose one small PDF bundle whose documents you recognize.
Start analysis with automatic confirmation off.
Preview each group, fix its name and pages, and confirm.
Open the resulting PDFs and check that all pages are present.
Download the results or send them into workspace processing. Check Documents for their next steps.

Edit, reset, and cancel carefully

Use a result’s name editor to change the proposed output name and page assignment before confirmation.
Add a manual group when the analysis missed a separate document.
Reset restores the editable drafts offered by the current review workflow; check the resulting groups before confirming.
During processing, the available cancel control may target the current item or a broader queue. Read its label before confirming.
Cancellation does not mean that outputs from previously completed jobs were deleted.

Missing preview on a narrow screen

The PDF preview can be hidden when there is not enough room beside the result controls. Use a wider window or the available file-opening workflow when you need to inspect pages. Keep reviewing the group boundaries before confirmation.

How does PDF Split work?