Privacy policy.

This Privacy Policy explains how Candy Studio Apps (“we”, “us”) handles personal data when you use Wisper AI for Android, including the floating bubble, widgets and Quick Settings tile (“Wisper” or the “App”). Wisper is a voice dictation app: you speak, and it turns your speech into clean text. We wrote this policy to be specific about what leaves your device, where it goes and for how long.

Separate policies apply to Wisper on iPhone, iPad and Mac — see the links at the top of this page.

1. Who we are

The controller responsible for your personal data is:

You can contact us about anything in this policy, including requests to exercise your rights, at support@wisperai.app.

2. What we collect

Wisper works without an account. You never give us your name, phone number or password. Here is what Wisper does process:

2.1 Content you create

• Voice recordings — the audio you record when you dictate with the floating bubble or record a note in the app.
• Transcripts and notes — the text produced from your recordings, plus summaries, key points, flashcards, rewrites and note chat that you ask for.
• Text you choose to edit — when you use Edit with voice, the text you selected and your spoken instruction.
• Your settings — dictionary words, snippets, voice styles, custom style descriptions, languages and translation targets.

2.2 Identifiers

• A random user ID created by our subscription provider, RevenueCat. It is not linked to your name or email.
• Analytics device IDs — random IDs that our analytics and crash-reporting tools assign to your installation, and Firebase installation IDs.
• Android Advertising ID — read by Meta’s SDK for ad measurement, unless you have deleted it or opted out in your Android settings (see section 5.4).

2.3 Usage and device data

• App events — for example that a dictation started or finished, how long it was, word and character counts, which features and styles you use, paywall views, and error codes. These events don’t contain the words you said or typed.
• Device and app information — device model, operating system version, app version, language, region, and microphone input type.
• Destination app — the name and package of the app you dictated into with the bubble (for example “WhatsApp”, com.whatsapp). See section 5.2.
• Crash reports — technical information about crashes and app stability.
• Network data — your IP address is processed by our servers and service providers when the App connects to them.

2.4 Purchases

If you subscribe, Google Play processes your payment. We never receive your card details. Through RevenueCat we receive your subscription status, plan, price, currency, purchase and renewal dates and store transaction IDs.

2.5 Feedback and support

• In-app feedback — when you send a feature request or tell us what to improve, we receive your message (up to 1,000 characters) and the email address if you choose to add one.
• Emails — if you write to us, we receive your email address and message. The “Contact us” email the App prepares includes technical details such as your device model, OS version and user ID so we can help you; you can remove them before sending.

3. What we don’t collect

• We don’t read passwords or other content on your screen, and we don’t record your screen.
• We don’t record audio when you aren’t dictating or recording.
• We don’t create voiceprints or use your voice to identify you.
• We don’t collect your contacts, photos, location, health data or browsing history.
• We don’t sell your personal data, and we don’t use what you say or write for advertising.

4. How your voice is processed

4.1 Dictation and recordings

1. You start recording. Tap the bubble or the record button. The microphone turns on only after your tap.
2. Audio is sent for transcription. Unless you use offline mode, your audio is sent over an encrypted connection (TLS) to our servers, which pass it to a speech-to-text provider. Depending on the settings, audio is sent as a file after you finish or streamed while you speak.
3. Text is cleaned up. The raw transcript is sent to a language-model provider together with what is needed to apply your choices: your style or template, dictionary words, snippet trigger phrases, language and translation target.
4. The text comes back to your device, where it is inserted where your cursor is and saved in your history.

Offline mode. If you download the on-device speech model (about 640 MB, downloaded from the sherpa-onnx project on GitHub) and choose Private mode, your audio is transcribed on your phone and is not sent to us. In Smart mode, Wisper uses the cloud when you’re online and the on-device model when you’re offline. Edit with voice always needs our servers.

4.2 AI features

When you ask for a summary, key points, flashcards, a rewrite or a note chat answer, the text of that note (and, for chat, your recent messages in that chat) is sent to our servers and a language-model provider to produce the result. When you use Edit with voice, the selected text (up to 5,000 characters) and your spoken instruction are sent the same way.

4.3 What our servers keep

• Audio: not stored. It is held in memory only while it is being transcribed.
• Transcripts: the finished text is kept for up to 30 minutes so the App can collect it if the connection drops, then deleted automatically.
• Edit results: the result of an edit is kept for up to 24 hours so a retried request doesn’t run twice, then deleted automatically. The text you selected is not stored.
• AI feature requests (summaries, flashcards, rewrites, chat): not stored.
• Logs: our servers log technical details such as your user ID, request times, audio length and the AI model used — not the words you said.

4.4 Our AI providers

We currently use these providers to transcribe and process your content. Each one processes it on our behalf only to return the result:

• OpenAI — speech-to-text, text cleanup, and AI features.
• Cerebras — fast text cleanup and edits.
• Anthropic — backup for AI features if our primary model is unavailable.
• ElevenLabs, xAI and OpenRouter — alternative speech-to-text providers we may route audio to for quality or reliability.

We use these providers through their business APIs. Under those terms they do not use your audio or text to train their models. Some providers may keep request data for a limited time — typically up to 30 days — to detect abuse, and then delete it. We don’t use your recordings or text to train AI models either.

5. Device permissions

Wisper asks only for the permissions it needs. You can change them at any time in your device settings; features that depend on a permission stop working without it.

5.1 Microphone and recording

Used to record your dictations and notes. The microphone turns on only after you tap the bubble or the record button, never in the background on its own. While recording, Wisper runs a foreground service with a visible notification, which lets an in-app recording continue with the screen off for up to 45 minutes.

5.2 Accessibility Service

Wisper uses Android’s Accessibility Service to provide dictation in other apps. Wisper is not an accessibility tool for people with disabilities; it uses this permission only for the following, after showing you an in-app explanation and asking for your consent:

• Showing the bubble at the right time — it detects when an on-screen keyboard opens or closes and which app is in the foreground, so the bubble can appear over text fields.
• Inserting your text — after you finish a dictation, it inserts the text into the field you’re typing in, by setting the field’s text, pasting from the clipboard, or committing it through the keyboard. When it pastes, it puts your previous clipboard content back afterwards.
• Reading the focused field — only at the moment of insertion, it reads the text already in that field so your dictation is added rather than replacing it.
• Edit with voice — if you turn this feature on, it reads the text you select in the focused field. That text is sent to our servers only when you dictate an edit instruction.
• Recording the destination app — it notes the name and package of the app you dictated into (for example “WhatsApp”, com.whatsapp). This is saved in your history on your device to show app icons, filters and Insights, and is included in analytics events about bubble use so we can see which apps Wisper works well with. If you turn on automatic style by app, it is also sent with your dictation so the right style is applied.

The Accessibility Service skips password fields, does not record your screen, and does not read, store or send other content you see or type in other apps. Its data is never used for advertising and never sold. You can turn it off any time in Settings → Accessibility; the bubble then stops working, and you can still dictate inside the Wisper app.

5.3 Floating bubble, notifications and widgets

The “display over other apps” permission lets Wisper show the floating bubble. You can switch the bubble off in Wisper’s settings. If you allow notifications, Wisper may send local reminders (for example a streak reminder or a weekly recap) and push messages about new features and offers through Firebase Cloud Messaging; turn them off in Settings → Push notifications or in Android settings. Home-screen widgets can show your latest dictation, so anyone who can see your home screen can read it.

5.4 Advertising ID and ad measurement

We use Meta’s SDK to measure how our ads perform. It receives app events such as installs, app opens and purchases, together with your Android Advertising ID if available, and Google Play’s install referrer. We never share your recordings, transcripts or notes with advertising partners. You can reset or delete your Advertising ID in Android Settings → Privacy → Ads (or Google → Ads). Meta processes this data as an independent controller under its own privacy policy.

6. How we use data and our legal bases

We process personal data only for the purposes below. If you are in the European Economic Area or the UK, the GDPR requires a legal basis for each:

We don’t make decisions about you based solely on automated processing that have legal or similarly significant effects.

7. Service providers and sharing

We share personal data only with service providers that process it on our behalf under contracts requiring them to protect it and use it only to provide their service to us, and with the parties described below.

We may also disclose personal data when required by law or a valid request from a public authority; to protect the rights, safety and security of our users, us or others; or as part of a merger, acquisition or sale of assets, in which case we will tell you before your data becomes subject to a different privacy policy.

Google Play handles your payments under Google’s own privacy policy; we don’t control it.

8. International transfers

We are based in the European Union. Several of our providers are in the United States or process data there. When personal data is transferred outside the European Economic Area, we rely on the European Commission’s Standard Contractual Clauses included in our providers’ data processing terms, or on the EU–US Data Privacy Framework for certified providers, together with safeguards such as encryption in transit. You can ask us for a copy of the relevant safeguards at support@wisperai.app.

9. How long we keep data

10. Security

All data between the App, our servers and our providers is encrypted in transit with HTTPS/TLS. We don’t store your audio on our servers, and we keep transcripts there only for minutes. Access to our production systems is limited to people who need it. On your device, your notes are protected by Android’s app sandboxing and your device passcode. No method of transmission or storage is completely secure; if a personal data breach is likely to put your rights at risk, we will notify the supervisory authority within 72 hours and inform you where the law requires it.

11. Your choices and deleting data

• Delete a note: delete it in the App. Its text and audio are removed from your device.
• Delete everything on your device: delete the Wisper app, or clear its storage in Android Settings → Apps → Wisper. If Android backup is on, a copy of the app’s data may also be in your Google account backup, which you control in your Android or Google account settings.
• Delete data held by us: email support@wisperai.app from the App’s Contact us option or include your User ID (Settings → User ID). We will delete your user record, usage counts and the analytics data linked to your user ID, except purchase records we must keep by law.
• Export your notes: use Export in Settings to save a file with your notes that you can keep or import on another device.
• Withdraw consent: revoke the microphone, Accessibility, overlay or notification permission in your device settings. Withdrawal doesn’t affect processing that took place before it.
• Process speech only on your phone: download the offline model and switch to Private mode.

12. Your rights (EU and UK)

Under the GDPR you have the right to:

• access your personal data and get a copy;
• rectify inaccurate data;
• erase your data (“right to be forgotten”);
• restrict processing;
• data portability — receive the data you gave us in a structured, machine-readable format;
• object to processing based on legitimate interests, including ad measurement;
• withdraw consent at any time, without affecting earlier processing.

Because Wisper has no accounts, we identify your data by your user ID; please include it in your request (Settings → User ID). We answer within one month; for complex requests we may extend this by up to two further months and will tell you why.

You also have the right to lodge a complaint with a supervisory authority, in particular in the EU country where you live or work. Our lead authority is the President of the Personal Data Protection Office (Prezes Urzędu Ochrony Danych Osobowych, UODO), ul. Stawki 2, 00-193 Warszawa, Poland, uodo.gov.pl.

13. US state privacy rights

If you live in California, Colorado, Connecticut, Virginia, Utah, Texas, Oregon or another US state with a consumer privacy law, you have the right to know what personal information we collect and how we use and disclose it; to access, correct and delete it; to receive a portable copy; and not to be treated differently for exercising these rights.

• Categories we collect (sections 2 and 6): identifiers (user ID, device IDs, advertising ID, IP address); audio information (voice recordings); commercial information (subscription purchases); internet or other electronic activity (app usage events); and your content (transcripts, notes, selected text). We collect them from you and your device, and use them for the purposes in section 6.
• Sale and sharing. We don’t sell personal information for money. Using Meta’s SDK to measure our ads (section 5.4) may count as “sharing” or “targeted advertising” under some state laws. You can opt out by deleting your Advertising ID in Android settings, or by emailing us — we will treat a Global Privacy Control signal sent with your request as an opt-out.
• Sensitive information. Voice recordings are used only to provide the service you request. We don’t use them to identify you, to infer characteristics about you, or for any purpose that would require a right to limit.
• Retention is described in section 9. In the past 12 months we have not sold personal information or knowingly processed data of anyone under 16.

To make a request, email support@wisperai.app with your user ID. You may use an authorized agent; we may ask for proof of the authorization. If we decline your request, you can appeal by replying to our decision; if the appeal is denied, you can contact your state attorney general.

14. Children

Wisper is not directed at children. You must be at least 16 to use it, or the age of digital consent in your country if lower (and never under 13), unless a parent or guardian has agreed. We don’t knowingly collect personal data from children under 13. If you believe a child has given us personal data, contact us and we will delete it.

15. Changes to this policy

We will update this policy when Wisper or the law changes. The date at the top shows the latest version. If a change is significant, we will tell you in the App or through the Google Play listing before it takes effect.

16. Contact

Questions, requests or complaints about privacy: