How to flatten a PDF - and when to do it

Last reviewed 2026-05-04. "Flattening" a PDF is the operation of baking the visible state of every layered object - form fields, annotations, comments, signatures, stamps, layers - into the static page graphics, so the result looks identical to a viewer but no longer carries the editable / re-fillable / re-extractable structure underneath. The free Flatten PDF tool runs the operation server-side and returns the flattened file for download. Most readers reach for it for one of three reasons: an e-signing service rejected a PDF that still had live form fields, a court / audit system expects a static archival copy, or a downstream tool keeps re-rendering an annotation that should have stayed put.

What flattening actually does to a PDF

A PDF (the page-description format defined by ISO 32000-1 / ISO 32000-2) is a tree of layered objects. The visible page is one stream of drawing operators; on top of that the file can carry an AcroForm dictionary (fillable form fields), an array of annotations per page (comments, stamps, sticky notes, free-text, ink markups, redaction marks), an optional signature dictionary (the cryptographic signature plus its visible appearance), and optional Optional Content Groups (layers that can be toggled on or off). All of these sit "above" the static page like transparent overlays. A flatten operation walks this tree and re-paints each overlay into the underlying page content stream as plain ink: the form-field value becomes static text drawn at the field's position, the annotation becomes a fixed shape with a fixed colour, the signature's visible appearance becomes a static image of the signature glyph, and the layer-toggle structure is collapsed to whatever was visible at flatten time. After flattening, opening the PDF in any reader shows the same pixels - but the AcroForm dictionary, annotation arrays, and Optional Content Groups have been removed, so there is nothing left to fill, edit, or toggle.

When you actually need to flatten

Flattening solves a concrete category of problem: the PDF needs to render identically everywhere and stay non-interactive. The four common situations:

• E-signing services that reject layered PDFs. Many e-sign platforms refuse to apply a final cryptographic signature on top of a PDF that still has live form fields or unsigned signature placeholders, because the form values can change after signing. Flattening the form first turns the field values into static text, after which the e-sign step can wrap the whole document in one cryptographic envelope.
• Court / audit / regulatory submissions. Filings often specify a static archival format - sometimes PDF/A, sometimes "flattened PDF". The reviewer needs the assurance that the file they received is exactly what every other reviewer will see; layered objects and form fields create the possibility of a different rendered page on a different reader.
• Sharing annotated drafts as final. When the editorial review is finished and the marked-up PDF should ship as the final, flattening bakes the comments and highlights into the page so they cannot be deleted by a downstream reader who clicks the wrong button. The annotations remain visible as ink rather than as removable overlays.
• Locking a signed page so the visible signature cannot drift. A visible signature appearance that lives only as an annotation can be moved, hidden, or removed by an editor. Flattening converts the appearance to static page content; the cryptographic signature dictionary is a separate concern, but the visible glyph stops being movable.

When NOT to flatten

Flattening is one-way. The original layered structure is gone after the flatten step; you cannot un-flatten a baked PDF back into a fillable form. Three cases where flattening is the wrong move:

• Downstream readers still need to fill the form. If the PDF is a tax form, a job application, or any blank template that the next reader is supposed to type into, flattening it before sending leaves them with a static picture of empty boxes. Send the layered file; let the recipient fill, then ask them to flatten before returning if the workflow needs an unalterable copy.
• A future editor expects to keep editing. Layered annotations and comments are how reviewers communicate. Flattening a draft erases the dialog - the next reviewer cannot delete a stale comment because it is no longer a comment, just ink on the page. Keep the layered version for the editorial loop and only flatten the final.
• Tagged-PDF accessibility depends on logical structure. A tagged PDF carries a structure tree (paragraphs, headings, list items, table cells) that screen readers traverse. Some flatten implementations preserve tags; others collapse them. If the document is consumed by assistive technology, verify the tags survive the flatten step, or do not flatten at all.

Flattening on freetoolonline

Open Flatten PDF, upload the PDF, and download the flattened result. The file is processed on the freetoolonline AWS service: it is uploaded via the standard upload pipeline, the flatten operation runs server-side, and the output file is rotated off the service storage after the retention window so you should download and save it locally if you need to keep it. The tool returns one PDF where every form field, annotation, and visible signature appearance has been baked into the page; the file size will usually drop slightly (the editable structure is gone) and the result opens with the same pixels in any PDF reader.

What flattening is NOT

Three operations are often confused with flattening; each is a different action with a different result.

• "Save as PDF" from a print dialog. This re-renders the PDF through a virtual printer, which usually does flatten layers - but it can also re-rasterise vector content (so text becomes a picture of text and stops being selectable / searchable). Flatten preserves the underlying drawing operators wherever possible; print-to-PDF may not.
• Compressing a PDF. Compression rewrites the page content streams to take fewer bytes - it does NOT remove the AcroForm dictionary, annotation arrays, or layers. A compressed PDF is still editable; a flattened PDF is not. See PDF Preflight Online - What It Actually Checks for the structural-checking counterpart that is also distinct from flattening.
• Removing a password. Password removal changes the encryption / permissions dictionary; it does not touch the layered objects. A password-removed PDF is still fillable / editable. See PDF Password Types - Owner vs User for the two password types and How to Sign a PDF After Removing a Password for the sign-after-remove workflow.

Companion guides on the PDF cluster: PDF Editing Ladder (the ladder of editing options before flatten is the right step), PDF Preflight Online - What It Actually Checks (the print-readiness check distinct from flatten), and PDF vs HEIC for Document Archival (format choice for the archival copy that flatten produces).