Privacy Policy
No legal jargon maze. No 47-page wall of text designed to confuse you into agreeing to something terrible. Just a plain breakdown of what we do and don't do with your data. Last updated: September 29, 2026.
What We Do vs. What We Don't
If you read nothing else on this page, read this.
✓ What We DO
- ✓ Process your text to generate audio (then drop it, unless you turn history on)
- ✓ Keep a usage record of each generation (voice, length, time) to apply your plan limits and stop abuse
- ✓ Save your dark/light theme preference locally
- ✓ Log basic server traffic for security (IP, page, time)
- ✓ Use Cloudflare for DDoS protection and delivery
- ✓ Encrypt all connections via HTTPS/TLS
- ✓ Respond to legal requests if compelled by law
✗ What We DON'T
- ✗ Require an account to use the free tool
- ✗ Keep your text when history is off (which is the default)
- ✗ Sell your personal data to anyone
- ✗ Use your text to train AI models
- ✗ Inject audio ads into your generated speech
- ✗ Share data with brokers outside our service providers
What Happens to Your Text
Exactly what happens from the moment you hit Generate to the moment your file is deleted.
You Hit Generate
Encrypted HTTPS request sent with your text and voice settings.
Server Processes
FastAPI receives text. No logging. Passes directly to TTS engine.
Audio Generated
MP3 saved to temp folder with random UUID filename. Inaccessible to others.
You Download
Browser fetches and plays your audio. You download the file.
Auto-Deleted
Background task deletes all files older than 1 hour. Automatically. No exceptions.
Your Rights Under Privacy Laws
GDPR, CCPA, and similar laws give you rights over your personal data. Here's how each one applies to FreeTTS.
Right to Know
What data do we collect about you?
Right to Delete
Can you request deletion of your data?
Right to Opt Out of Sale
Can you stop us from selling your data?
Right to Portability
Can you export your data?
Right to Non-Discrimination
Will we treat you differently if you exercise rights?
GDPR & CCPA Compliance
For users in the EU, California, and anywhere else with strong privacy laws.
🇪🇺 GDPR (European Users)
- Our servers are in Germany (Hetzner, Falkenstein). Speech, moderation and database providers process outside the EU (US and Singapore) under their standard contractual clauses.
- Legal basis: your explicit consent when you click Generate.
- Data minimization: we collect the absolute minimum needed.
- Auto-deletion within 1 hour exercises your right to erasure automatically.
- No formal DPO required given our minimal processing scope.
- Contact [email protected] for any data protection inquiries.
🇺🇸 CCPA (California Users)
- Right to Know: detailed in this policy. Spoiler: almost nothing collected.
- Right to Delete: auto-deletion handles this before you'd request it.
- Right to Opt Out of Sale: we don't sell personal information. Period.
- No “Do Not Sell My Personal Information” link because there's nothing to sell.
- Non-discrimination: everyone gets the same free service regardless.
- Categories collected in last 12 months: only standard server logs.
Complete Privacy Policy
The full details, still written in plain English. No legal maze.
What Information We Collect
FreeTTS does not require you to create an account, provide an email address, or hand over any personal information. You can use the entire tool without telling us a single thing about yourself.
If you contact us at [email protected], we'll have your email from that exchange. We use it only to respond. We don't add it to a mailing list or share it with anyone.
When you convert text, that text is sent to our server for processing. It's received, used to generate audio, and automatically deleted within one hour. We don't log it, read it, or analyze it.
Like every website, our server logs basic connection info: IP address, page requested, browser type, and visit time. These are standard security logs, rotated regularly, not analyzed to identify individual users.
Cookies and Similar Technologies
FreeTTS uses essential cookies, plus the analytics and advertising cookies described below:
Essential cookies on the website. These are required for the site to function. They remember your dark/light mode preference, your authentication session if you sign in to a PRO account, and security tokens that protect against cross-site request forgery. You cannot disable essential cookies without breaking core functionality.
First-visit cookie. On your first page view we store a small first-party cookie (ft_trace, 30 days) and a matching browser-storage entry holding the page you landed on, the site that linked you here if any, and any utm tags in the address. If you later create an account, that is saved with it so we can see which sources bring people who actually use FreeTTS. It holds a path and a hostname, no personal data, and it is not shared with anyone.
Affiliate link cookie.If you arrive through an affiliate partner's link (an address ending in ?via= and a name), we keep that name and the time of the click in a first-party cookie (ft_via, 90 days) so the partner is credited if you buy. Nothing about your browsing is recorded, and it is only read when you pay. If you do buy, your payment record notes which partner referred it; the partner sees the plan and the amount of the sale in their dashboard, never your name or email address.
Affiliate partners. If you join the affiliate program, we keep your name, email address, the username you choose and your referred sales, to run the program and pay you. If you save payout details on your Affiliate page (a Wise email, a bank account for ACH or wire, or a USDT wallet address), they are stored encrypted, shown only to you and to the FreeTTS owner, and used only to send you your commissions. You can change them at any time; they are deleted with your account.
Analytics cookies. The FreeTTS website uses Google Analytics to understand aggregate, largely anonymized usage (which pages are visited, broad device and region) so we can improve the product. It is not used to build advertising profiles of you.
Advertising cookies (Google AdSense).The website may display ads served by Google AdSense to free users. Third-party vendors, including Google, use cookies to serve ads based on your prior visits to this and other websites, and Google’s use of advertising cookies enables it and its partners to serve personalized ads. You can opt out of personalized advertising in Google Ads Settings, review how Google uses data from sites that use its services at policies.google.com/technologies/partner-sites, and opt out of additional third-party vendors at aboutads.info. The text you convert, your uploaded files, and your generated audio are never shared with advertisers or used to target ads. PRO and Creator subscribers see an ad-free experience.
Browser extension storage. The FreeTTS Chrome extension does not use cookies at all. It stores preferences (selected voice, playback speed, install ID, listen queue, recent generations, optional auth token if you sign in for PRO) locally via chrome.storage.local. This data never leaves your device except for the explicit API calls described in the “Browser Extension” section below.
If you do not wish to accept any cookies at all, most browsers allow you to block cookies entirely through your browser settings. Note that doing so may prevent parts of FreeTTS from working correctly, particularly the account sign-in flow.
Where your data is processed
Our own servers are in the EU. Most of the specialist services we depend on are not, so here is the full list rather than a summary. Where processing happens outside the EU or EEA, it is covered by the standard contractual clauses each provider publishes.
- FreeTTS servers (the website and API) — Hetzner, Falkenstein, Germany (EU).
- Supabase (accounts, usage records, saved projects, and history text if you switch it on) — Singapore.
- Microsoft Azure Speech (most voices, and Speech to Text) — United States, with a European region used only as a late fallback.
- Google Cloud Text-to-Speech (Signature voices) — United States.
- ElevenLabs (Ultra voices) — United States.
- OpenAI (safety screening of submitted text, AI Summary, caption timing for videos with HD or Signature voices, and the Studio's co-direct and subtitle translation) — United States.
- Mistral AI (reading the pages of a scanned PDF for a PRO audiobook) — European Union.
- Google Gemini (choosing stock footage for faceless videos, and pronunciation suggestions in the Studio dictionary) — United States.
- Pexels and Pixabay (stock footage search for faceless videos; they receive search words only) — stock libraries run by Canva.
- Cloudflare (delivery and abuse protection) — a global edge network; your request is served by the location nearest you.
- Stripe and Freemius (payments, and PayPal for subscriptions bought before 18 September 2026) — each is its own controller for payment data and processes it under its own privacy policy, which may involve transfers outside the EU.
- Apple and Google Play (purchases made in the FreeTTS app on iPhone and Android) — each is its own controller for payment data under its own privacy policy; we receive the purchase and the plan, never card details.
- Resend (transactional email) — United States.
In plain terms: text you submit for speech leaves the EU. It goes to OpenAI for the safety check and then to the provider that owns the voice you chose, and those calls are served mainly from the United States. If that is not acceptable for your material, please do not submit it.
Third-Party Services
Cloudflare— We use Cloudflare as our content delivery network and security layer. Cloudflare may process your IP address and request data as part of delivering the site and protecting it from abuse. Cloudflare has its own privacy policy governing this processing.
Speech providers — Microsoft Azure, Google Cloud and ElevenLabs— Text that you submit for speech synthesis is forwarded to whichever of these provides the voice you selected, only to generate the audio. Each processes the text solely to return audio and, under their enterprise terms, does not retain it or reuse it for training. Microsoft Azure also turns recordings and audio files into text for Speech to Text; the recording is deleted once the transcript is made.
Mistral AI— When a PRO user turns a scanned PDF into an audiobook (a PDF with no readable text layer), the PDF is sent to Mistral's OCR service to read its pages. Word files and PDFs that already contain text never go to Mistral. Mistral hosts this data in the European Union.
Google Gemini — Used for two things. Faceless videos:when you choose the automatic stock footage background, each scene's spoken line is sent to Gemini to write a stock footage search, and the candidate clips' thumbnails are sent to it to pick the best match. Pronunciation suggestions:in the Studio's pronunciation dictionary (PRO), a word and its language are sent to Gemini when our offline dictionary cannot suggest a pronunciation. No user identifier is sent.
Pexels and Pixabay— With the automatic stock footage background, the search words for each scene are sent to these stock libraries to find clips. No account data is sent.
Supabase— If you choose to create a PRO account, authentication and account data are handled by Supabase under their privacy policy.
Stripe & Freemius— Paid checkout and subscription billing run on two rails. Card payments through our Stripe checkout are processed by Stripe, which acts as the merchant of record for those purchases and is responsible for charging and remitting any sales tax or VAT. Purchases made through the Freemius checkout, including when you choose to pay with PayPal there, are sold by Freemius, Inc., which acts as the merchant of record for those purchases and handles your payment details under its own privacy policy. Subscriptions bought directly through PayPal before 18 September 2026 continue to be billed by PayPal until they end. Purchases made before July 2026 were processed by Creem. FreeTTS never sees or stores your card details on any of these rails.
FreeTTS MCP server— If you add https://mcp.freetts.org/mcp to an AI assistant, the text the assistant sends is handled like text typed on this website and goes to the same providers. Calls without an account are metered by the connection's IP address. If you connect your account, the server stores the API key you paste, on our servers, to act for you; delete the key in the dashboard to disconnect. The server logs the tool name, plan, character count and outcome, never the text.
Resend— Transactional emails (welcome, payment confirmation, password reset) are delivered through Resend.
OpenAI — Safety screening:text you submit for speech generation is sent to OpenAI's moderation API before it is synthesised, so we can refuse content our speech providers and our payment partners prohibit. This applies to every generation route — the free tool, PRO and Creator generation, multi-voice dialogue, voice previews, the public API, audiobook batches and video voiceovers. Text longer than 8,000 characters is screened in slices. AI Summary: if you invoke that feature, the text you ask to summarise is forwarded to OpenAI to produce the summary. Caption timing:for faceless videos and added voiceovers read by an HD or Signature voice, the generated audio and the start of the script are sent to OpenAI's transcription model so captions land on the right words.Studio tools on the website:co-direct sends the lines of a dialogue for delivery suggestions, and subtitle translation sends the subtitle text. In every case no user identifier (email, account ID, or IP) is sent with the request, and per OpenAI's API data-usage policy, API inputs are not used to train OpenAI models. Neither the submitted text nor the summary is retained by FreeTTS after delivery. If a request is refused, we record the policy category, the score, the time, the endpoint and the character count — never the text itself.
Data Sharing
FreeTTS does not sell your personal information. The only parties that may receive your data are the third-party service providers listed above, each of which processes data solely to provide a service to FreeTTS (hosting, delivery, speech synthesis, speech to text, AI summarisation, reading scanned PDFs, choosing stock footage, payment processing, authentication, email) and operates under its own privacy policy.
We may also disclose information if legally required to do so by a valid court order, subpoena, or law enforcement request. Given how little account data we collect, the amount we would be able to provide in response to such a request is minimal.
Data Security
- All connections encrypted via HTTPS/TLS
- Industry standard server security configuration
- Temporary audio files stored in restricted directory, auto-deleted
- We don't store sensitive personal info because we don't collect it
- Security headers: HSTS, CSP, X-Frame-Options, Referrer-Policy
The most effective data security is not collecting data you don't need. We follow that principle seriously.
Children's Privacy
FreeTTS does not knowingly collect personal information from anyone, including children under 13. Since we don't collect personal information from anyone (no accounts, no signups), the tool is safe for users of all ages. Parents: your kids can use this safely.
International Users
FreeTTS is used by people in over 100 countries. We apply the same privacy standards to everyone, regardless of location. EU users don't get better privacy protections than users in Asia or Africa. Everyone gets the same minimal data treatment. If your country has specific regulations not addressed here, email us and we'll work with you.
Browser Extension (FreeTTS for Chrome)
The FreeTTS Chrome extension is covered by this same privacy policy. This section spells out exactly what the extension does with your data, separately from the website.
What the extension does on your device. The extension stores the following locally via chrome.storage.localand never syncs anything across devices: your preferred voice, playback speed, most recent text, a randomly generated install ID, a 24-hour audio cache (so repeated text plays instantly), the Listen Queue (saved articles you want to hear later), recent AI Summary results, and — only if you sign in on freetts.org — a copy of your Supabase auth token and resolved plan (free / PRO / Creator / Lifetime). Uninstalling the extension clears all of this data automatically.
What leaves your device. The extension only sends data to the network when you explicitly invoke a Read, Theatre Mode, Summarise, or Download action. In those cases:
- Read / Theatre Mode / Download: the text you selected (or the article body of the page you right-clicked) is sent to
freetts.org/api/ttsfor speech synthesis. As with every generation route, it is screened by OpenAI's moderation API first (no user identifier attached). The text is processed in memory, the resulting MP3 is returned, and both are deleted within one hour. - AI Summary (PRO): the text you asked to summarise is sent to
freetts.org/api/summaryand forwarded server-side to OpenAI to generate the summary. No user identifier (email, account ID, or IP) is sent to OpenAI with the text. The summary is returned to you and not retained. - Optional sign-in: if you choose to sign in at freetts.org, the extension reads your Supabase session token from the freetts.org page and sends it to
freetts.org/api/auth/meto verify your plan and unlock PRO features. The token is sent only to FreeTTS, never to any third party.
What the extension does NOT do. No advertising. No analytics. No behavioral tracking. No tracking pixels. No Google Analytics, AdSense, or Mixpanel. No reading of pages other than the one you explicitly target via right-click or keyboard shortcut. No background scanning of your browsing history or open tabs. No data sold to anyone.
Crash reporting. The extension includes Sentry to report its own crashes — a stack trace, the extension version, and which part of the extension failed. It never sends page content, page URLs, the text you convert, or your audio. On Chrome and Edge it is on by defaultand you can switch it off at any time in the extension's Settings under Privacy. On Firefox it is off by default and only runs if you turn it on.
What host access the extension asks for. The manifest declares one host permission: https://freetts.org/*, so the extension can reach our own API. To read a page, it uses activeTab together with scripting, which grant access to a single tab only at the moment you invoke the extension — via right-click, the toolbar button, or a keyboard shortcut — and not before. That access is what lets the service worker fetch audio on your behalf so playback still works on sites with a strict Content Security Policy, such as Wikipedia or banking portals, and what lets the Listen Queue open and read a saved article in a new tab. The extension never reads or transmits page contents in the background.
Audio output. The extension never injects audio advertisements into the speech it generates, and never modifies, sells, or transmits the audio it plays anywhere other than back to you.
Changes to This Policy
If we change this policy, we'll update the “last updated” date at the top. For significant changes, we'll make it obvious on the website. We're not going to sneak changes past you. That would be a crummy thing to do.
Contact
Questions? Concerns? Want to verify we practice what we preach? Email us: [email protected]. We respond within 24 hours. A real person writes back. Every time.
The FreeTTS App: iPhone, iPad and Android
The app (org.freetts.app) is available for iPhone and iPad on the App Store and for Android on Google Play. Both use the same account and the same servers as this website. This is what it sends and keeps.
What the app sends to our servers
- The text you ask it to read, or the address of a web page you ask it to read, so we can turn it into speech. Handled exactly like text typed on this website.
- The name of a voice when you tap to hear its preview, so we can play the sample.
- Speech to Text: the microphone records only while you hold or tap record. The recording, or a file you pick, is sent to us for the transcript and then deleted.
- Add a voiceover: the video you pick is uploaded only to render it with the new audio. Your video stays on our servers.
- A PDF or Word file you pick for a book, uploaded so we can split it into chapters.
- Before the app sends anything to a third party AI service, it asks your permission and names each one: OpenAI (safety check of text, and caption timing), Microsoft Azure (voices and Speech to Text), Google (Signature voices, and Gemini for automatic stock footage), ElevenLabs (Ultra voices) and Mistral (scanned PDFs). Faceless videos with automatic stock footage also send search words to Pexels and Pixabay. Details are under “Where is my data stored?” above.
- Your sign-in (Apple, Google or an email code) through Supabase, our account provider.
- A signed purchase receipt from Apple when you buy a plan, so we can turn your plan on. Apple bills you; we never see your card.
- On Android, Google Play billing: Google is its own controller for your payment data. We receive the purchase token and the plan you bought, which we check with Google to turn your plan on, and never your card details.
- Crash and failure reports to Sentry (EU servers, Germany): what went wrong, the device model, the app build, and for a failed request the address path without ids, the status and our error code. Never the text you read, a recording, your email, your name or your account id.
- An install id (a random number made once on your device) on free, signed-out use, so free limits apply per device rather than per shared network address.
What stays on your device, and deletion
- Audio you download, your recordings, rendered videos and subtitle files are kept on your device until you delete them, and are removed with the app. Your history, favorites and settings follow your account, as on the website.
- The app is for people aged 13 and over.
- Delete your account in the app: Account → Delete account and recordings. This removes your account, your generation history, favorites, saved books, transcripts and recordings, and revokes Sign in with Apple. If you have an App Store subscription, cancel it in your Apple ID settings, and a Google Play subscription in the Play Store under Payments and subscriptions; deleting the account does not cancel Apple or Google billing. You can also delete your account without the app: see Delete your account.
- Purchases made in the app are managed and refunded by Apple, or by Google Play on Android, under their own terms. See Terms, App purchases.
Common Privacy Questions
No. FreeTTS does not require any signup or personal information to use. No name, no email, no phone number, nothing. You can use the entire tool completely anonymously. The only data that touches our server is the text you convert, and that gets deleted within one hour.
No, by default. Your text is sent to our server solely to generate the audio file. Both the text (in memory) and the temporary audio file are automatically and permanently deleted within one hour via a background cleanup task. We do not log, read, analyze, share, or use your text for AI training.
Optional history (signed-in users only): if you opt in via Dashboard → Privacy & data → "Save my generation history", we keep a short metadata row per generation (the first 500 characters of your text, voice, timestamp, and your IP) so you can replay items from the History tab. The toggle is OFF by default. You can flip it back at any time, and you can delete your existing history in one click from the same screen. We never enable history without your explicit opt-in.
Essential cookies (theme preference, your PRO sign-in session, anti-CSRF security tokens). The website also uses Google Analytics and may show Google AdSense ads to free users — see the Cookies section below for details and opt-out links; the text you convert is never used for ads, and PRO is ad-free. The browser extension does not use cookies at all — it stores its preferences locally via chrome.storage.local.
Our own servers are located in Germany (Hetzner, Falkenstein), and we apply GDPR protections to everyone regardless of where they are. Being straight about one thing: FreeTTS is not a closed EU system. To turn your text into speech we send it to specialist providers — Microsoft Azure, Google and ElevenLabs — and to OpenAI's moderation API for the safety check, and those run mainly in the United States. Our database (Supabase) is hosted in Singapore. So your text and account data are transferred outside the EU, under the standard contractual clauses each of those providers publishes. We collect minimal data, delete free and guest audio within about an hour, and answer data protection requests at [email protected].
No. FreeTTS does not sell your personal data. We work with standard third-party service providers — Cloudflare (delivery), Microsoft Azure, Google Cloud and ElevenLabs (speech, and Azure for speech to text), Supabase (accounts), Stripe and Freemius (payments; PayPal only for subscriptions bought before 18 September 2026), Apple and Google Play (purchases in the FreeTTS app), Resend (email), OpenAI (safety screening, AI Summary, caption timing and subtitle tools), Mistral (reading scanned PDFs), Google Gemini (choosing stock footage and pronunciation suggestions) and Pexels and Pixabay (stock footage search); details below. Each operates under its own privacy policy and processes data only as needed to provide its service to FreeTTS. We do not sell customer lists, share personal information with data brokers, or send your text anywhere except to the services named on this page, for the feature you use.
Our own servers are hosted by Hetzner in Falkenstein, Germany, and every request reaches us there first. That is not the whole picture, though, so here it is in full. Your account, usage records and saved projects live in our database with Supabase, hosted in Singapore. Text you submit for speech is sent to OpenAI's moderation API for the safety check, then to whichever speech provider owns the voice you picked — Microsoft Azure, Google or ElevenLabs — and those calls are served mainly from the United States. Each provider processes it only to return the result, under the standard contractual clauses they publish for international transfers. Audio for free and guest generations is deleted about an hour after it is made.