Privacy policy
What stays local.What leaves the browser.
DocTranslate AI parses documents in the browser first, but translation still involves sending extracted text through external services.
This page explains that model in plain language so users can decide whether it fits the sensitivity of their document.
What is handled in the browser.
The service is designed so the first stage of document handling happens locally in the browser.
When you choose a supported file, the app first reads and parses it locally in your browser to extract the text needed for translation or verification.
For certified PDF verification, the check is performed locally in your browser by extracting text from the file and recomputing the relevant hash.
The public translation flow does not currently require account creation, profile setup, or a personal dashboard.
What may be sent to external services.
Translation still requires extracted text to pass through external services.
To translate a document, extracted text segments may be sent over HTTPS to external services involved in bot protection, request handling, and language generation.
These services can include Cloudflare Turnstile, Supabase-hosted functions, and the third-party model provider used to generate translations.
Because of that, this workflow should not be treated as fully offline or as a confidential vault for highly sensitive records.
What this site does not guarantee.
This page explains how the public service works, without making guarantees that depend on third-party providers.
This site does not promise that infrastructure providers or translation providers operate without their own logging, retention, or compliance obligations.
This site is not presented as a secure document vault, a long-term repository, or a privacy product for privileged, regulated, or highly sensitive material.
This site does not promise error-free translation, uninterrupted availability, or suitability for legal, medical, financial, or regulatory reliance without human review.
Contact and policy updates.
Questions about the public behavior of the service can be directed to the project contact listed on the site.
For questions about this policy or the public-facing behavior of DocTranslate AI, contact rpn_doc@proton.me.
This policy may be updated as the product, providers, compliance requirements, or site monetization setup evolve.
If advertising or consent tooling is added later, the privacy disclosures on this site should be updated to reflect those changes before rollout.
Common questions.
These answers restate the public boundaries of this page in plain language.
Does the original document file leave my device?
The workflow starts locally in the browser, where the file is parsed first. Translation still requires extracted text segments to be sent to external services over HTTPS.
Is this a fully offline translator?
No. It is browser-first, not fully offline. Parsing and rebuilding start locally, but translation still depends on external request handling, bot protection, and model inference.
Should highly sensitive material be treated as safe by default here?
No. If a document is sensitive enough that external processing is unacceptable, this workflow should not be treated as a confidential vault for that file.
Does DocTranslate AI promise what third-party providers do with their own systems?
No. This site does not promise that infrastructure or translation providers operate without their own logging, retention, or compliance obligations.
Need the usage boundaries?Read the terms.
The privacy page explains handling. The terms explain responsibility, acceptable use, and what PDF certification does not mean.