What Bindery is
Bindery provides 29 browser workflows for organizing, compressing, converting, OCR, signing, protecting, redacting, repairing and automating PDFs. The app is designed for focused document jobs rather than full paragraph-level PDF editing or cloud collaboration. Ordinary use requires no account, and the installable web app can continue core workflows offline after its required assets have loaded.
The product is operated through binderypdf.com. Questions about accounts, billing, privacy or corrections can be sent to [email protected].
The processing boundary
PDF engines run inside the browser tab. When a user selects a file, the browser reads it into local memory, creates the result and offers that result for saving. Bindery does not provide an upload endpoint for document processing.
This does not mean the website makes no network requests. Cloudflare delivers and protects the site and provides cookieless page analytics. Optional accounts use Supabase for identity, usage totals, subscription state and synced recipe definitions. Stripe handles payment details. Recipe records describe a sequence of tool steps; they do not contain the PDFs run through that recipe. The full boundary is documented in the privacy policy.
How product behavior is tested
Tests cover wiring, security boundaries, billing entitlements, usage metering, generated metadata, canonical URLs, language alternates and sitemap completeness. Functional checks use real PDF bytes rather than only checking that a button exists.
- Merges and page extraction are reopened and checked for the expected page count and order.
- Password workflows use both wrong and correct passwords and verify whether the resulting file remains encrypted.
- Form tests create actual fields, fill them, flatten the output and verify that no editable fields remain.
- Metadata tests write values, read them back and confirm that stripping removes the intended fields.
- OCR checks use known text in supported languages and compare recognized output, including accented characters.
- Compression and repair outputs are parsed again so a smaller or rewritten file is not treated as success merely because bytes were downloaded.
Production releases also receive live HTTP, structured-data, responsive-browser and Lighthouse checks. A green automated suite is evidence for covered behavior, not a claim that every possible PDF is valid or recoverable.
How guides and comparisons are researched
Technical guides distinguish architecture from marketing language and explain trade-offs such as compression quality, OCR error, redaction flattening and device memory. Comparison pages are dated and link to the competing service’s own security, pricing or product documentation for claims that can change.
Comparisons deliberately include a section describing where the established product is stronger. Bindery does not treat every cloud workflow as unsafe or pretend that focused browser tools replace enterprise administration, certificate-backed signatures, collaborative storage or advanced content editing.
Corrections and update policy
Product limits and competitor policies change. Material comparison claims display an update date and should be checked against their linked primary source before a high-stakes decision. If a factual statement is outdated or an output behaves differently from its documentation, email [email protected] with the page URL, browser and a non-sensitive reproduction. Never send a confidential source PDF for a public-content correction.
Corrections are made in the page generator as well as the published HTML so they persist through future releases. Sitemap and IndexNow submissions are refreshed when new canonical pages are published.
What Bindery does not claim
- Local processing does not protect a compromised device or malicious browser extension.
- No imposed file-size paywall does not mean a device has infinite memory.
- OCR is probabilistic and important names, totals and identifiers need human verification.
- A drawn page signature is not a certificate-backed digital signature.
- Repair can rebuild parseable PDF structure but cannot recreate missing bytes.
- PDF-to-Word conversion cannot perfectly reconstruct every fixed-layout design.