How can we help?

Drop files anywhere on the home page, or pick a tool from the catalog, the Tools menu or the ⌘K search. PDFs open in Organize (several PDFs open Merge), images open Images to PDF, and Office or text documents open Any file to PDF.

Every tool has the same shape: files on the left, options on the right, one primary button. Results show a preview with Download, Save to My documents and “Continue with” shortcuts into the next tool.

Ranges refer to the original document: 1, 3-5 selects page 1 and pages 3 to 5. Leave the field empty to include every page. In Split, Extract and Remove you can also click page thumbnails (Shift-click selects a run of pages).

Downloads go to your browser’s download folder; several files arrive as one ZIP. “Save to My documents” keeps a copy in this browser’s storage so you can reopen it in another tool later. Nothing is synchronised between devices.

Use Unlock PDF with the password you know; it removes protection from a new copy, locally. Protect PDF adds an AES-256 opening password. Passwords are used only for the current operation and never stored or uploaded.

Lossless compression rewrites the structure, so already-optimised PDFs stay the same size — we keep the original rather than make it larger. Balanced and Smallest re-render pages as images: much smaller for scans, but selectable text, links and forms are lost.

Edit PDF and Add text place new content on top of the page; existing text is not rewritten. Crop hides margins but keeps the content in the file. Redact removes selected regions by rebuilding every page as a flattened image — searchable text, forms and links are gone from the result, so review every page before sharing.

No. It places a drawn or typed visual signature on the page. It does not create a cryptographic signature, verify identity or provide an audit trail.

OCR recognises English text in scanned pages on your device and produces a searchable PDF plus a text file, up to 20 pages per task. Recognition can make mistakes; check the result before relying on it.

A workflow always starts with Merge, then runs the steps you saved — compress, rotate, crop, watermark, page numbers — with the settings you chose. Open Workflows, build or pick a template, then press Run: the Merge tool opens with your workflow attached.

PDF/A-2b (ISO 19005-2), the archival profile: an embedded sRGB output intent, XMP identification, a permanent document ID, and no encryption, scripts or attachments. The tool first reports what will change. If every font is embedded and no CMYK colour is used, the content is kept as-is and stays searchable; otherwise pages are rasterized to sRGB images, which is always conformant but no longer selectable (run OCR afterwards if you need search).

Structure is checked on your device. For a formal archive submission, confirm the file with an independent validator such as veraPDF.

Into PDF: Word (.docx), Excel (.xlsx), PowerPoint (.pptx), OpenDocument (.odt, .ods, .odp), RTF, EPUB, HTML, Markdown, plain text, CSV, HEIC photos and common images.

Out of PDF: Word, Excel, PowerPoint, PNG, JPG, WebP, Markdown and text. Plus CSV to Excel and HEIC to JPG.

Legacy binary formats (.doc, .xls, .ppt) and Apple iWork files must be saved as their modern equivalents first.

Styles, headings, lists, tables with merged cells, pictures, footnotes, hyperlinks, page size and margins, headers and footers all come through. Headings become PDF bookmarks and links stay clickable.

Charts, SmartArt, macros, tracked-change markup and embedded objects are skipped and listed on the result screen. Fonts are substituted with Noto (sans, serif, mono) with automatic fallback for Arabic, Hebrew, Devanagari, Thai and CJK text.

Text positions are read from the PDF and rebuilt into headings, paragraphs, lists and tables; running headers and footers are removed; scanned pages are kept as pictures. The result is a clean, editable document rather than a pixel-exact copy. For image-only PDFs, run OCR first if you need editable text.

The converter never fails a whole document for one unsupported feature. Anything it had to leave out — an OLE object, a chart, a metafile picture — is listed under the result so you know exactly what to check.

Device-first. Browser built-in AI (Chrome/Edge) and the on-device WebLLM model run entirely on your computer and cost nothing. The cloud engine is included in paid plans and used only when you pick it and confirm; it receives extracted text, never the file. You can also bring your own provider key.

The model (roughly 1–2.6 GB depending on your choice) is downloaded once and cached by your browser. It needs WebGPU, so a recent desktop browser with a GPU works best. Remove it any time from Settings → AI.

No. PDF tools, conversions and on-device AI run inside your browser using WebAssembly and workers. There is no upload step, no server-side queue and no retention window. Only account, billing and (if you allow it) anonymous page statistics reach our servers.

Page views and active visitors, counted only after you allow it and never when your browser sends Do Not Track or Global Privacy Control. No document data, no cross-site tracking, and you can switch it off in Settings at any time.

Saved results live in this browser’s storage. They are not uploaded or synchronised, so clearing site data removes them. Download anything you want to keep.

No. All 49 local tools work as a guest with the Free limits (50 MB per file, 200 pages per task). An account keeps favorites and workflows per person and is where a subscription attaches once billing opens.

Create a key on the Developers page, then POST a multipart request with your file to /api/v1/convert (to=pdf|docx|xlsx|pptx) or /api/v1/pdf/{operation}. The response is the file itself. Each successful operation uses one credit; failed calls are free. API files are processed in memory on the server and never stored — unlike the web app, which never uploads at all.

Larger files and batches, more saved workflows, an ad-free workspace and the cloud AI engine. Every local tool and conversion is the same on every plan.

Choose “Forgot password?” in the sign-in dialog: we email a one-hour link to set a new password, and every other device is signed out once you do. To change your email, open Settings → Account → Change; the new address must confirm a link before sign-in moves to it.

Settings → Security lists every signed-in device with a last-active time, lets you sign any of them out and shows a log of recent account changes.

Every tool works before you verify. Verification proves the address is yours so a password reset can reach you, and it is required to create a team, because team owners manage other people’s access. Resend the link from Settings → Account any time.

A Business owner creates a team and invites colleagues by email; invitations last seven days and are tied to the exact address. Members inherit the owner’s Business entitlements — larger files, cloud AI, no ads — while the owner’s plan is active. Owners handle billing, ownership transfer and deletion; admins manage members and invitations.

Teams share a plan, never files: documents still stay on each person’s own device. The activity log records membership and role changes only.

Yes, both from Settings → Privacy & data. Export downloads a JSON file with your profile, teams, devices, security log and API usage. Deletion is immediate and removes sessions, API keys and team memberships; teams you own must be transferred or emptied first. Documents were never stored by us, so there is nothing else to remove.