Kebijakan Privasi

Overview

Your files are uploaded to our transcription servers. Audio is deleted after processing; unfinished uploads and queued jobs expire after 24 hours. Results can be recovered for seven days with a private key saved in this browser. You can delete them earlier. An internet connection is required. Limited diagnostic metadata, without audio or transcript content, is retained for up to seven days after a job changes state, including after deletion.

This policy explains what limited product, analytics, advertising, and browser-stored data may still be involved when you use the site.

What data we process

Audio and transcription content

Your files are uploaded to our transcription servers. Audio is deleted after processing; unfinished uploads and queued jobs expire after 24 hours. Results can be recovered for seven days with a private key saved in this browser. You can delete them earlier. An internet connection is required. Limited diagnostic metadata, without audio or transcript content, is retained for up to seven days after a job changes state, including after deletion.

Audio fragments are sent securely through Hugging Face Inference Providers to Deep Infra Inc. for transcription. Hugging Face states that it does not store routed request bodies or responses, and retains debugging logs without user content or tokens for up to 30 days. DeepInfra states that it operates its inference infrastructure in the United States; its terms provide for deletion of request content after processing and prohibit using it to train models, subject to their exceptions. Our separate job recovery and diagnostic retention periods still apply. Routing information: https://huggingface.co/docs/inference-providers/en/security; provider terms: https://deepinfra.com/terms; privacy: https://deepinfra.com/privacy.

Browser-stored data

Recovery keys and transcripts are stored in this browser until you delete them or clear site storage. Anyone with access to this browser profile may access them. Download a backup before clearing storage. Audio files are not saved in browser history.

Usage and reliability data

Our transcription server records job states, upload and queue timings, processing duration, cancellation, and reported result delivery and export actions to operate and diagnose the service. These records exclude audio, transcript text, filenames, IP addresses and advertising identifiers. Private job diagnostics expire seven days after their last state change; statistics aggregated by day, without job identifiers, are kept for up to 90 days. Browser observations do not prove that a result was read or that its text was accurate.

For consumption accounting we retain content-free provider request receipts for up to 35 days, including attempted audio duration, the reference rate and success or uncertainty. Daily totals without request or job identifiers are retained for up to 90 days. Estimates are kept separate from billed amounts.

Only when you grant analytics permission, newly submitted jobs may also be grouped using a private, salted version of this browser’s existing service identifier to measure repeat use, completed minutes and simulated free allowances over 7 and 28 days. These optional records expire 35 days after job creation. They are not sent to Google or advertisers and cannot identify the same person across devices. Withdrawing analytics permission removes this browser’s optional grouping when a saved job receipt can reach our server; offline requests retry when you return. Clearing browser storage removes the saved receipts needed for this request, and remaining records expire automatically. Transcription remains available without analytics permission.

Advertising data

We use Google AdSense to support the free product. Depending on your consent choice and Google’s policies, Google and other advertising partners may use cookies, web beacons, and similar identifiers to collect or infer data such as page context, browser/device details, IP address (or IP-derived region), advertising identifiers, and ad interaction events.

For details, see How Google uses data when you use our partners’ sites or apps.

How we use data

• Your files are uploaded to our transcription servers. Audio is deleted after processing; unfinished uploads and queued jobs expire after 24 hours. Results can be recovered for seven days with a private key saved in this browser. You can delete them earlier. An internet connection is required. Limited diagnostic metadata, without audio or transcript content, is retained for up to seven days after a job changes state, including after deletion.
• Transkripsi online
• To monitor errors, performance, and product reliability
• To measure aggregate product usage
• To serve ads that help keep TalkToTextly free
• To respond to support requests you send us

Third-party services

• Google AdSense — advertising
• Sentry — error monitoring and debugging
• Online transcription
• Your audio is securely uploaded for transcription.
• InMobi — consent management
• Google Analytics 4 — optional, consent-based analytics
• Audio fragments are sent securely through Hugging Face Inference Providers to Deep Infra Inc. for transcription. Hugging Face states that it does not store routed request bodies or responses, and retains debugging logs without user content or tokens for up to 30 days. DeepInfra states that it operates its inference infrastructure in the United States; its terms provide for deletion of request content after processing and prohibit using it to train models, subject to their exceptions. Our separate job recovery and diagnostic retention periods still apply. Routing information: https://huggingface.co/docs/inference-providers/en/security; provider terms: https://deepinfra.com/terms; privacy: https://deepinfra.com/privacy.
• Optional email sign-in uses AWS Cognito. Stripe hosts any available pack checkout and processes payment details; we do not store card numbers. Private account, job and payment references are not sent to Google analytics or advertising.

Cookies and consent

We use InMobi’s certified consent platform worldwide. You can accept, reject or manage purposes and advertising partners, and reopen your choices using the privacy button at the bottom of the page. Accepting can enable personalized ads. If you decline, Google may serve limited ads without personalization; these may use storage solely to prevent fraud and IP addresses to deliver ads. Ad availability is not guaranteed. Declining does not restrict transcription.

InMobi processes your choices and technical information, such as IP-derived region and browser details, to display the appropriate notice and communicate your preferences to advertising providers. It saves consent preferences in browser storage, currently for up to 13 months. Audio files and transcript text are not sent to InMobi, AdSense or Google Analytics.

Google Analytics 4 loads only after you grant its separate analytics permission in the InMobi privacy choices. It measures page visits, referring sites and content-free tool events, using analytics cookies. It does not receive audio, transcript text, filenames, job identifiers or recovery links. We disable Google signals and ad personalization in Analytics. You can decline or withdraw analytics consent without restricting transcription or changing your advertising choices. Events that occurred before permission are not replayed. Our server continues to measure service reliability independently. URL queries and fragments are omitted from Analytics; campaign attribution is therefore limited.

To provide the free audio allowance, we use an essential secure browser cookie for up to 35 days and keep quota usage records for up to 35 days. Short-lived protected network identifiers help prevent repeated anonymous use and are normally removed after 24 hours. These functional records do not depend on optional analytics consent. Clearing the cookie does not guarantee a new allowance.

Optional account sessions use a secure, HttpOnly cookie for up to one hour; sign-in state expires after 10 minutes. Signing in links the browser’s existing free usage to the account rather than creating a new free allowance.

Security

• HTTPS is used for the site and third-party requests
• Audio Anda diunggah dengan aman untuk ditranskripsikan.
• No account is required to use the core tool
• We minimize telemetry and avoid sending audio or file names in analytics events
• Encrypted recovery snapshots can retain protected account and quota references for up to 120 days under the current backup lifecycle. They exclude audio, transcripts, filenames and sign-in session tokens. Purchase records and unused credit have a separate accounting and support lifecycle; closing access does not erase those records or automatically delete the managed sign-in profile. Contact support for access or deletion requests.

Your choices

• Reject personalized advertising and manage your privacy preferences
• Clear browser storage to remove local history and preferences
• Read the support page for practical file, browser, mobile, export, and troubleshooting guidance
• Contact us with privacy questions or deletion requests for support emails you sent us directly