Privacy Policy

1. Who we are

Levelang is a translation service operated by Alex Beverage. If you have questions about this policy or your data, contact us at privacy@levelang.app.

2. What data we collect

We collect only what is necessary to operate the service:

• Account information — your email address, display name, password, and any preferences you configure (such as preferred languages and translation settings) when you create an account.
• Sign-in with Google — if you choose to sign in with Google, we receive your name, email address, and profile photo from Google.
• Device information — when you use the mobile app, we collect a device identifier, platform (iOS or Android), and app version to authenticate your device and enforce usage limits. Device identifiers are platform-assigned UUIDs, not hardware serial numbers or advertising IDs.
• Translation content — text you submit for translation (and audio, if you use voice translation) is sent to a third-party AI provider for processing. We do not store the text you submit or your voice-translation recordings — they are processed in real time and discarded. If you play a translation aloud, see Spoken translations below.
• Chat messages — what you write in the website's chat is sent to Google's Gemini to generate the reply. We do not store your messages; we keep only a count of how many you send each day, to apply the daily limit.
• Your voice (optional) — if voices are enabled for your account and you record yourself on the voice page, we send the recording (10 to 60 seconds) to our voice provider, Gradium, which makes a version of your voice in each language you choose. We do not store the recording. For each voice you keep, we store its identifier at Gradium, its language and accent, the language you recorded in, and when you agreed to the voice terms and which version you agreed to. Playback in your voice is AI-generated speech, made by Gradium from your recording.
• Spoken translations — when you play a translation aloud, its text is sent to Gradium to generate the audio, in your own voice or a standard one. We keep the generated audio and its words for up to 14 days after it was last played, so a repeat plays at once, and delete it as soon as you remove the voice it was made with. We also store your preferred playback speed.
• Waitlist — your email address if you join our waitlist before creating an account.
• Request logs — like any website, our hosting providers record technical details of each request (your IP address, browser type and the page requested) to deliver it and keep the service secure. We don't run analytics or tracking scripts.
• Operational telemetry — the mobile app sends performance traces (request timing, device model, OS version) to help us monitor service health. These traces do not contain translation content or personal information.
• Error reports — automatic crash reports via Sentry to help us fix bugs. These may include your browser type and the page URL, but no personal information.

We do not collect payment information, location data, contacts, or any special-category personal data. Voices made from your recording are used only to speak translations to you, never to identify you.

3. Why we collect it (legal basis)

Under the GDPR, we process your data based on the following legal grounds:

• Contract performance — when you create an account or register your device, we process your data as necessary to provide the translation service you signed up for, including authentication, storing your preferences, enforcing usage limits, and managing your account.
• Consent — you provide your email address voluntarily when joining the waitlist. You can withdraw consent at any time.
• Consent (your voice) — we make and use voices from your recording only with your explicit consent, given on the voice page each time you record. You can withdraw it at any time by removing your voices there or by deleting your account.
• Legitimate interest — request logs, operational telemetry, and error reports help us secure, maintain and improve the service.

4. How we use your data

• To authenticate you and keep you signed in.
• To provide the translation service, including sending your text or audio to an AI provider for processing.
• If you choose to, to make versions of your voice from your recording and to speak translations to you in them.
• To store your preferences for a personalised experience.
• To enforce usage limits and prevent abuse.
• To send you updates about Levelang (launch announcements, early access invitations).
• To maintain and improve the website and service.
• To diagnose and fix technical issues.

We will never sell your personal data or share it with third parties for marketing purposes.

5. Where your data is stored

Account and device data is stored in Supabase (hosted in the EU). If you sign in with Google, your identity is verified by Google and subject to their privacy policy.

Translation text and audio are sent to Google Gemini for processing. Google processes this data under their API terms of service — translation content sent via the API is not used to train their models. We do not store your translation content, apart from the audio of translations you play aloud (below). Chat messages are sent to Google's Gemini models on Vertex AI, processed in the EU (Belgium).

If you use your own voice, your recording and the voices made from it are processed by Gradium, our voice provider, which stores voices in the EU. Text you play aloud is sent to Gradium to generate speech on its nearest servers, in the EU or the United States. Gradium processes all of this for us under a data processing agreement: it may not use it to train its models or for any purpose of its own, and keeps it only while it provides the service to us. We keep generated audio in Supabase (EU) for up to 14 days.

The website and our servers run on Fly.io, in Paris. Sign-in and waitlist emails are sent through Resend. Our domain's DNS is managed by Cloudflare, which also redirects levelang.app to www.levelang.app. These providers see technical details such as your IP address in order to deliver each request or email.

Operational telemetry is sent to Grafana Cloud for service monitoring. Error reports are processed by Sentry.

Some of these providers are based in, or may process data in, the United States: Google (Gemini), Gradium (speech generation), Fly.io, Resend, Cloudflare, Grafana Cloud and Sentry. Those transfers rely on the EU–US Data Privacy Framework or on Standard Contractual Clauses approved by the European Commission.

The mobile app stores your preferences and device registration locally on your device. All providers maintain GDPR-compliant data processing agreements.

6. How long we keep it

We retain your data for as long as you have an active account or remain on the waitlist. Device registrations are automatically deleted after 180 days of inactivity.

Voices made from your recording expire after 30 days. The temporary copy Gradium makes from the recording is deleted as soon as your voices are ready, and voices you don't choose are deleted within 90 minutes. You can remove your voices at any time on the voice page: we delete them at Gradium straight away, and keep retrying until that succeeds. Generated audio is deleted 14 days after it was last played, or as soon as its voice is removed.

You can delete your account at any time from the app or website — this permanently removes your data from our systems via cascading deletion, and deletes your voices at Gradium. If you request erasure by email instead, we will remove your personal data within 30 days.

7. Your rights

Under the GDPR, you have the right to:

• Access — request a copy of the data we hold about you.
• Rectification — ask us to correct inaccurate data.
• Erasure — ask us to delete your data ("right to be forgotten").
• Portability — receive your data in a machine-readable format.
• Restriction — ask us to temporarily stop processing your data while a concern is resolved.
• Object — object to processing based on legitimate interest. We will stop unless we have compelling grounds to continue.
• Withdraw consent — remove your voices on the voice page at any time, and unsubscribe from emails using the link in any email we send, or by contacting us directly.
• Directions after death — under French law, you can tell us what should happen to your data after your death.
• Lodge a complaint — with the CNIL, the French data protection authority, or with the authority where you live, if you believe we have not handled your data correctly.

To exercise any of these rights, email privacy@levelang.app. We will respond within 30 days.

8. Children's privacy

Levelang is open to users of all ages. If you are under 16 years old, you must have a parent or guardian's permission before providing your email address or any personal data. If we learn that we have collected personal data from a child under 16 without parental consent, we will delete that data promptly. If you believe we may hold data about a child without proper consent, please contact us at privacy@levelang.app.

9. Cookies and browser storage

We do not use cookies for tracking or advertising. Essential cookies may be set by our hosting infrastructure for security purposes only. When you sign in, your browser stores session data to keep you signed in between visits. The website also remembers, in your browser, which voice you chose for each language. The voice demo at voice.levelang.app sets a single cookie to remember that you entered its passcode. None of these are used for tracking.

The mobile app uses on-device storage to save your preferences (language, level, theme) and device registration. This data stays on your device and is not sent to us unless required for the translation service. You can clear this data from the app settings.

10. Mobile app permissions

The Levelang mobile app requests microphone access only if you use voice translation. Audio is recorded locally, sent to our server for translation, and immediately discarded — it is never stored. No other device permissions (camera, contacts, location) are requested.

On the website, only the voice page asks for microphone access, to record your voice as described above. The recording is sent to Gradium and is not stored by us.

11. Changes to this policy

We may update this policy from time to time. If we make significant changes, we will notify you via email (if we have your address) or by posting a notice on the site. The "last updated" date at the top reflects the most recent revision.