Privacy Policy
How ScribeText handles your account, recordings, transcripts and privacy choices.
Last updated: September 15, 2026
About this policy
ScribeText is an audio-to-text service at scribetext.com. In this policy, “ScribeText”, “we” and “us” mean the team operating that service. For privacy questions or requests, email hi@scribetext.com.
This policy covers the ScribeText website and its account features. A recording can contain information about people other than the account holder. Please provide the necessary notices and obtain any permission required before submitting someone else's personal information.
Recordings and the sample workspace
Choosing a file first creates a local playback preview. Clicking Start transcription uploads that file to private Cloudflare R2 storage and starts processing after account, format and available-minute checks. Soniox receives access to that recording through a temporary link so it can create a transcript. The link is sent by our server to the provider and is not a public file listing.
Your transcript, edits and requested AI notes are saved with your account. When you request notes, transcript text is sent to Cloudflare Workers AI to produce a summary and a basic mind map. Long transcripts are processed in sections and then combined.
The separate sample workspace uses illustrative content. Its sample edits are stored in this browser until you reset the sample or clear site data. Downloaded exports and recordings in your device's folders remain under your control.
Information we handle
Depending on the features you use, information can include:
- Account information: your name, email address, profile image, authentication information and account settings. If you choose Google sign-in when available, we receive the profile and authentication information needed to sign you in, not your Google password.
- Service and security information: account and session identifiers, IP address, browser information, access times and records needed to operate the service and investigate errors or misuse.
- Information you submit: support messages and any files you explicitly upload through an available upload feature, together with filenames, file type, size and ownership information.
- Recording feedback: when you report a problem from a recording, we associate your message with your account and recording, and save task status, provider request identifiers, error codes, usage information, the current view and playback position, and a basic browser label. You can optionally share a short excerpt from the saved transcript. Feedback does not attach your audio or full transcript. Support notification emails contain a report reference and an administrator link, not your message or transcript excerpt.
- Payment information, when paid plans are available: customer, subscription, checkout and invoice identifiers, plan and payment status, and billing-period records. Stripe handles payment details in its checkout; ScribeText does not receive or store your full card number or card security code through that checkout.
- Recordings and results: the submitted media, spoken content, language choices, timestamps, speaker labels, transcript edits and requested summaries or mind maps. These can contain personal or sensitive information, so submit only what you have authority to process.
Why we use information
We use information to provide the features you request, maintain accounts, respond to support requests, manage purchases when available, secure the service and meet applicable legal obligations.
Where data protection law requires a legal basis, account and requested-service processing generally relates to providing the service; proportionate security and troubleshooting relate to legitimate interests; legally required records relate to legal obligations. Processing that requires consent must have a separate valid consent mechanism. Reading this policy is not consent to optional tracking, marketing or unrelated use of your recordings.
ScribeText does not sell recordings or transcripts. We do not use them to train general-purpose AI models. This statement does not replace a provider's own data-handling terms; applicable provider terms must be checked before that provider receives content.
Service providers and disclosure
We disclose information only as needed for the relevant service, your instructions, a lawful request, or protection against fraud and security threats.
| Provider or service | Purpose and current scope |
|---|---|
| Cloudflare | Hosting and private file storage, account and result data, account email delivery, and Workers AI for requested summaries. See Workers AI data usage. |
| Authentication if you choose Google sign-in and it is enabled. | |
| Stripe | Payment processing and subscription management when paid checkout is enabled. See Stripe's Privacy Policy. |
| Soniox | Speech-to-text processing for recordings you submit. Soniox states that API audio and transcripts are not used for model training. Its Async API stores submitted audio and results and currently automatically deletes them after 30 days, with earlier deletion available through its API. That is a provider policy, not a 30-day deletion promise for all ScribeText data. See Soniox's data-handling documentation. |
| Other transcription providers | The administrator configuration includes reserved integrations. A provider does not receive your recordings merely because its name appears in a settings page. |
A provider may process some information for its own legal, security or payment obligations. Its privacy notice explains those activities. The current service does not offer a customer-selected country of processing. Contact us before submitting content that requires a particular processing region or contractual arrangement.
Retention and deletion
- Local previews and sample edits: the browser behavior described above applies. Clearing local data does not delete a recording from your own folders or a file you downloaded.
- Account records: information needed to maintain your account is retained while the account is active. You can use the available account deletion control under Settings → Security or contact us. This does not by itself cancel a payment subscription or remove copies held separately by a payment processor.
- Support and security records: retention depends on the time needed to resolve your request, investigate an incident and meet applicable recordkeeping obligations. We review whether continued retention remains necessary.
- Recording feedback in the application: the report and any shared excerpt remain linked to the recording until you delete that recording or your account. Resolving a report does not delete it. Deleting only the original audio does not delete a feedback report associated with a saved transcript.
- Original cloud recordings: new uploads are scheduled for deletion 30 days after upload. A recording still being transcribed is protected until its job finishes. Failed deletion operations remain queued for retry. This rule does not retroactively assign a deletion date to older development uploads.
- Unfinished uploads: resumable upload sessions expire after 24 hours. Known incomplete parts are then queued for cleanup. If storage accepted an initialization request but its response was lost, its incomplete-upload lifecycle provides an additional cleanup safeguard.
- Transcripts, saved edits, summaries and mind maps: these remain with your account until you delete the associated record or your account. Deleting only the original audio through Files does not erase a saved transcript or restore used minutes.
- Provider copies: we request deletion of temporary Soniox resources after saving a result or closing a failed job. A job still processing or a request with an unconfirmed response may require later reconciliation. Provider deletion is asynchronous; it is not a promise of immediate removal from every provider or backup.
Deleting an account removes account-associated application records and queues the remaining known file and transcription-provider resources for deletion. Active subscriptions must first end through Billing or with support assistance. Minimal identifiers may be retained to finish resource cleanup and reconcile delayed payment events; unresolved payment events require operator review. Payment processors can retain records for their own obligations. Email us if your personal information appears in someone else's recording and you want to request its removal.
Cookies, local storage and communications
We use authentication cookies and local preferences where needed for account features and theme settings. Sample transcript edits and temporary upload-resume metadata are also stored in your browser. Details are in our Cookie Policy.
With your optional analytics choice, Google Analytics 4, our self-hosted Plausible instance at pla.panjin.dev, and our self-hosted Umami instance at umami.panjin.dev measure site visits and product steps such as selecting a recording, viewing results and exporting. Google receives browser/device information and an analytics cookie identifier. Plausible runs on infrastructure we manage through Zeabur, uses connection and browser information to produce aggregate statistics, and does not set analytics cookies. These are our own Plausible and Umami installations on Zeabur, not their managed cloud services. Umami uses connection and browser information for aggregate statistics, sets no analytics cookies, and respects Do Not Track. We send public tool types and export formats, use generic titles and aggregated private page paths, and remove query strings, fragments, filenames, account identifiers and recording content. Google and Plausible receive no referrer. Umami may receive a recognized public search or AI service origin without paths or queries; other referrers are discarded. Analytics is not used for advertising personalization or session replay. You can decline or withdraw analytics through Cookie preferences in the footer. See our Cookie Policy for storage, controls and provider information.
The optional support chat widget and product newsletter are not active. Account security messages and purchase receipts are separate from marketing.
Your choices and rights
Depending on where you live and the applicable law, you may have rights to access, correct, delete, obtain a portable copy of, restrict or object to processing of personal information, withdraw consent, or complain to a data protection authority. These rights can have conditions or exceptions.
Email hi@scribetext.com with your request. We may need to verify your identity and the information concerned before acting, without requesting more identification than necessary. We will respond within the period required by applicable law. Withdrawing consent does not affect processing that was lawful before withdrawal.
Security, children and updates
We use authentication and access controls to protect account features. No online service or device can guarantee absolute security. ScribeText is not designed for children; contact us if you believe a child has provided personal information without appropriate permission.
We will update this page when relevant practices change. Material changes will be brought to your attention as required, and any new consent that is legally necessary must be obtained before the affected processing starts.
For service-use rules, see our Terms of Service.