Privacy Policy
Last updated: July 7, 2026
/tools/ — run entirely in your web browser. The files and text you work on are never uploaded, never stored on our servers, and never seen by us. Our AI tools download their models to your device; nothing of yours is uploaded from it. Optional cloud-powered features, where offered, are clearly labeled before you use them — everything else runs on your device. There is nothing for us to delete because nothing ever leaves your device.
1. Who we are
SnipSound is a free website at snipsound.com that helps content creators preview music for their videos and provides a suite of free browser-based audio utilities. This policy explains what data we do and do not collect when you visit.
2. Files you process with our tools
When you open an audio or video file in any tool under snipsound.com/tools/ — including the Audio Trimmer, Audio & Video Converter, AI Transcription, Text-to-Speech, Voice Recorder, BPM Finder, Key Finder, and Audio Reverser — the file is decoded and processed using the Web Audio API, WebAssembly, and on-device AI models inside your own browser. The bytes of the file are never sent to any SnipSound server, third party, or cloud service. When you close the tab the file is released from browser memory.
Because there is no upload step, there is also:
- No file size limit imposed by us (your device’s available RAM is the only ceiling).
- No queue, no "delete after X hours" retention — there is nothing on our end to retain.
- No risk that your unreleased music, voice memos, or private recordings are sitting in a temp folder on someone else’s server.
Our AI tools download models to your device — not the other way around. Tools like AI Transcription (powered by Whisper) and Text-to-Speech (powered by Kokoro) fetch an open-source AI model file — roughly 75–310 MB depending on the tool and your device — from our servers the first time you use them, then run that model inside your browser. If you watch your network traffic you will see these large files downloading to you; you will never see your files going out. Both engines are Apache-licensed open-source models running in your own browser, where anyone can inspect exactly what the code does — you are trusting the architecture, not just our word.
A simple sanity check: a 2 GB video would take many minutes to upload on a typical home connection. Our tools start working on it in seconds — the upload you would have to sit through simply never happens.
Don’t take our word for it — verify it yourself
What lives on your device: the AI engine cache (so models don’t re-download), your preferences and settings, and — only if you use a save or resume feature — your working files. All of it sits in your browser’s own storage on your machine. Our servers receive none of it.
- The airplane test. Open a tool once so its engine downloads, then turn off Wi-Fi — and keep working. A tool that needed our servers would stop; ours don’t.
- The network tab. Open your browser’s developer tools and watch the traffic while you work. You will see: model files downloading to you, the site’s own assets, and the small analytics events described in section 4 (never file contents). You will never see your file’s bytes leaving your device.
- The disappearing act. Clear this site’s data in your browser and everything — models, settings, saved work — is gone. There is no server copy to restore from, because there was never a server copy.
3. Music files we serve to you
The music tracks you can preview on the Snip Editor and Playlist pages are hosted by SnipSound on Cloudflare. When you press play, your browser streams those files from our servers, the same way it streams any other web resource. We do not associate playback with your identity.
4. Information we collect automatically
Like most websites, our hosting provider logs basic, non-personally-identifying request data such as IP address, user-agent string, referring page, and the timestamp of each request. This is used to keep the service running, prevent abuse, and understand aggregate traffic patterns. Server logs are not linked to a SnipSound account because SnipSound has no account system.
Cookies and local storage
SnipSound uses your browser’s localStorage to remember preferences you set inside the app — for example, your liked tracks, hidden tracks, custom snippets, and language selection. This data lives on your own device and is not transmitted to us. You can clear it at any time by clearing your browser’s site data for snipsound.com.
We do not use third-party advertising cookies.
Analytics
We use PostHog to understand which pages and features people use, so we can prioritize what to improve. PostHog records aggregate page views, clicks, and basic browser metadata (browser version, operating system, country at the country level, referring URL). It does not collect names, email addresses, or the contents of any audio file you process — audio still never leaves your device. PostHog honors the browser’s Do Not Track setting; if you have DNT enabled, no analytics events are captured for your session. Text input values across the site are masked so anything you type (file names, FAQ search terms, etc.) is not sent. We do not use autocapture: only explicit product events (like “tool completed”) are sent — never arbitrary clicks, page content, or DOM data. PostHog data is hosted in the United States.
Session replay. PostHog also records visual replays of how some sessions interact with the page — mouse movements, scrolls, clicks, and which UI elements appeared. We use these recordings strictly to diagnose problems (e.g. “why did the export button freeze for this user?”) and improve the interface. All form inputs are masked, file names are not captured, and audio files never appear in a recording because they never leave your device in the first place. Sessions are not recorded if your browser has Do Not Track enabled.
Error reporting. Uncaught JavaScript errors that happen in your browser are automatically reported to PostHog so we can fix bugs — the report includes the error message and the page you were on, never any audio content or personal data.
5. Third-party services
We use the following third-party services strictly to deliver and improve the website:
- Cloudflare — hosts the website, audio files, and provides CDN and DDoS protection.
- PostHog — product analytics (pageviews and clicks only; no audio, no PII, DNT-respecting; US-hosted).
If you click through to an external site — for example a Spotify, Instagram, or TikTok link for one of our artists — that site’s own privacy policy applies once you arrive there.
6. SnipSound in ChatGPT and other AI assistants
SnipSound provides a music connector (an MCP server at mcp.snipsound.com) that lets AI assistants such as ChatGPT and Claude search our free-music library. When you use it:
- What we receive: only the search your assistant sends on your behalf — the query text and any filters (mood, genre, duration, energy). We do not receive your name, your account details, or any other part of your conversation.
- Nothing is logged or retained: the request is processed in memory to produce the search results and then discarded. The connector keeps no logs, sets no cookies, has no user accounts, and serves only public catalog information (track titles, genres, licenses, preview links).
- Previews and artwork: when you play a preview in the chat, the audio and images load directly from our media server (media.snipsound.com), like loading any web page. No analytics run inside the chat, and the in-chat player sends nothing back to us.
- Clicking through to SnipSound: links in the results carry standard attribution parameters (for example
utm_source=chatgpt) that tell us the visit came from an AI assistant. Once you are on snipsound.com, this privacy policy applies exactly as it does to any other visit.
If we ever add usage measurement to the connector (for example, anonymous aggregate counts of searches or previews), we will update this section before doing so.
7. Children
SnipSound is a general-audience tool. We do not knowingly collect personal information from anyone, including children.
8. Your rights
Since we do not collect personally identifiable information, there is generally no personal data for us to access, correct, or delete on your behalf. You retain full control of files processed in our tools because they remain on your device. To remove preferences stored locally, clear your browser’s site data for snipsound.com.
9. Changes to this policy
We may update this policy as the site evolves — for example, when new tools are added or new third-party services are introduced. The "Last updated" date at the top of this page will always reflect the latest revision. Continued use of the site after a change indicates acceptance of the updated policy.
10. Contact
Questions about this policy can be sent through the contact options on the SnipSound homepage.