Privacy Policy

Subtitle Translator is presented publicly as nosme, while paid checkout, receipts, and subscription records are operated by AI Master through Creem.

That means some billing metadata is held at the operator layer so the product can reconcile plan state, support requests, and payment events without hiding who is actually charging the customer.

During the public translator and signed-in workflow, Subtitle Translator processes subtitle files in memory, source character counts, file metadata, selected language pairs, and the minimum account data needed to track quota ownership. Uploaded subtitle files are not stored for later reuse.

The operational footprint stays narrow on purpose. We only collect the data required to estimate usage, route the job through the public provider path, and return a translated subtitle file that can still be reviewed before release.

2.1 Operational metadata

We retain only short-lived operational metadata such as job timestamps, selected target languages, character counts, and basic error states needed to keep the public workflow reviewable.

2.2 Subtitle job data

The live demo discards subtitle files after processing and export. Future paid workspaces may add short retention windows later, but those windows will only expand after the account system and policy copy stay aligned with the public product behavior.

• Source subtitle format and structural markers
• Selected target language pairs
• Source-character volume used for billing estimates
• Short-lived processing diagnostics for failed jobs

The current live product prefers Google Cloud Translation as the default processing provider because it is the most accessible route in the current operator setup while still offering a well-documented public API.

Subtitle Translator is an independent workflow product built around Google Cloud Translation. It is not the official Google product surface and does not claim affiliation beyond using that processing API.

Alternative routing remains disabled publicly until quality, fallback behavior, and policy wording are ready. We do not silently switch users onto a different provider path while the public copy still promises something else.

Privacy requests route through the live support channel for this product. Requests are reviewed against the current public behavior, including transient file processing, account-level quota tracking, and the absence of long-term file storage for subtitle uploads.