Start
Privacy and data handling
What leaves your device, what Scribiz keeps and for how long, what Google sees, and how to delete a result.
On this page
This page describes what happens to your media and your results. The privacy policy is the legal text. This page is the plain version.
What leaves your device
It depends on where you run Scribiz.
| You run | What is sent | To whom |
|---|---|---|
| YouTube captions: bulk batches, the YouTube transcript pages of the site, the caption tools of the MCP server, and any transcript paid with a caption credit | Public video, playlist and channel identifiers. A retrieval service returns metadata and existing caption text. No video or audio is sent to a model. | Scribiz and its caption retrieval service |
A YouTube transcript on your minutes, over MCP (get_transcript, search_video) or the API (mode: "auto" with detail: "transcript"): no caption credit left, or a video without captions | The caption text, which Gemini may punctuate when the captions are automatic and have no punctuation. A video without captions is read as in Listen, the next rows. | Scribiz, then the speech model or Google |
| Web, with a YouTube link for Listen or Watch | The link. YouTube blocks our server from downloading the video, so Google's model reads the public video from the link. No copy of the video is made. | Scribiz, then Google |
| Web, with another site's link or a direct media link | The link. Our server then downloads the audio, and for Watch a low-resolution copy, from the site. | Scribiz, then Google |
| Web, with a file | The audio. For a video file your browser extracts the audio first. | Scribiz, then the speech model named below |
| MCP media tools, with or without a key, a YouTube link | The link and your questions. Google's model reads the public video from the link. | Scribiz, then Google |
| CLI with your own Gemini key, Listen | The audio, as short 16 kHz mono chunks. Never the whole video. When signed in to Scribiz, the finished transcript and summary are saved to your account for History. | Google (your key), and Scribiz History when signed in |
| CLI with your own Gemini key, Watch or Both | A low-resolution copy of the video, built on your machine. When signed in to Scribiz, the finished result is saved to your account for History. | Google (your key), and Scribiz History when signed in |
| CLI with your own Gemini key, a public YouTube link read through Gemini | The link. Google's model reads the video from YouTube. When signed in to Scribiz, the finished result is saved to your account for History. | Google (your key), and Scribiz History when signed in |
| CLI signed in to Scribiz, Listen | The audio, as short chunks. Never the whole video. | Scribiz, then Google |
| CLI signed in to Scribiz, Watch or Both | A low-resolution copy of the video, built on your machine. | Scribiz, then Google |
| CLI or Mac app with your own OpenRouter key, Listen | The audio, as short chunks. When signed in to Scribiz, the finished transcript and summary are saved to your account for History. | OpenRouter (your key), and Scribiz History when signed in |
| CLI or Mac app with a local model | The audio is read on your computer by Whisper and never leaves it. When signed out, nothing leaves your computer; when signed in to Scribiz, the finished transcript is saved to your account for History. | Nobody when signed out; Scribiz History when signed in |
| CLI with browser cookies for a site | Nothing. Cookies are read locally and used by the local downloader. | Nobody |
The Mac app runs the CLI inside, so it sends the same things.
Scribiz never uploads a whole local video. For audio the file is cut into chunks on your machine. For video a small copy is made first.
The YouTube transcript pages of the site and the bulk caption routes retrieve existing text only. They do not listen, watch or generate summaries. Over MCP, a transcript that runs on your minutes is a media job: a video without captions is listened to. For a YouTube link explicitly submitted for media processing, Auto mode uses its captions first. YouTube blocks the download from our server, so to listen, Scribiz passes the link to Google and Google's model reads the public video. For other sites and direct media links, our server downloads the audio (and for Watch a low-resolution copy of the picture), sends it to Gemini and deletes it when the job ends. The transcript from a YouTube link read this way has approximate timing and no speaker labels.
What the hosted service keeps
- Media. Audio chunks and low-resolution copies are deleted from Google's file store when the run ends, success or failure. As a backstop, Scribiz removes any it missed after two hours, and Google expires files after at most 48 hours. A file you upload to the API is removed when its job succeeds. After a failure it is kept until its one hour is up, so a retry needs no second upload.
- Results. A result is kept so that you can come back to it. Results are kept for 24 hours, or for 30 days when you are signed in. You can delete one sooner. When you are signed in to a Scribiz account in the Mac app, the CLI or the local MCP server, finished transcripts and summaries (including runs processed with your own key or a local model, never the audio or video file) are saved to your account so they show in History across all your devices for 30 days, and you can delete them at any time. Signed-out use stays on your computer only.
- Caption batches. Bulk caption results and exports require the account or API key that created them. They are kept for 30 days. Download copies you want to keep.
- Private links. A single-video result lives at an unguessable address. It is not indexed by search engines and it is not in the sitemap. There are no public transcript pages.
- A reuse cache. To avoid paying twice for the same public video, Scribiz can reuse a stored result when someone else asks for the same video. The cache holds derived text, never media, and only for videos of a platform such as YouTube: never an upload and never a direct media link. Video details are kept for 24 hours. Successful raw YouTube caption snapshots are dated and kept for up to 365 days in a separate cache capped at 8 GiB. Other captions are kept for up to 30 days and generated text for up to 90 days in the existing 1 GB cache. Storage limits may evict entries earlier. Failed or partial retrievals are not stored as successful captions. It is never shown as a public page. On a takedown request it is purged.
What the speech model sees
When Scribiz reads audio for you (the web tool, the API, the chats, and the CLI or Mac app when you are signed in), the audio goes to ElevenLabs' Scribe v2 through OpenRouter, and to Google when that cannot answer. The audio travels inside the request and Scribiz keeps no file with either company. Scribiz asks OpenRouter to use only providers that do not keep what they are sent or train on it. That is their commitment under their own terms, and it can change. With your own Gemini key, your own OpenRouter key or a local model, the audio goes only where that choice says: see Speech models.
What Google sees
Summaries, chapters, the proofread and on-screen notes use Google's Gemini API, and so does a read of audio when you chose your own Gemini key or the speech model could not answer. A YouTube transcript paid with a caption credit, or made on the YouTube transcript pages of the site, is not sent to Gemini. One that runs on your minutes over MCP can be, to restore punctuation.
- The hosted service uses a Gemini API account with billing enabled. Google's Gemini API terms say that content sent with billing enabled is not used to improve Google's products. Read the terms for the exact wording.
- When you use your own key, Google's rules for that key apply. Keys without billing enabled may let Google use your content to improve its products. Turn on billing for the key to opt out.
Your keys and cookies
- An OpenRouter key goes to OpenRouter only. The CLI stores it in the same file with the same mode, or takes it from
OPENROUTER_API_KEYwhen you choose an OpenRouter model. The Mac app keeps it in the Keychain. - A Gemini key goes to Google only. The CLI stores it in
~/.scribiz/config.jsonwith mode 0600, and takes it from the environment if you setGEMINI_API_KEY. - The Scribiz key that
scribiz loginmakes goes to Scribiz only, and is stored in the same file with the same mode.scribiz logoutremoves it from that machine. It does not revoke it: revoke it in the dashboard. No flag takes a key, sopscannot show one. The Mac app, once it is out, will keep it in the Keychain. - Scribiz API keys are stored as a hash. Revoke a key in the dashboard and it stops working within about a minute.
- Browser cookies are used by the local downloader and are never sent to Scribiz. A hosted request for a site that needs a login fails and tells you to download the file and upload it.
Website analytics
- Analytics. The privacy policy says whether the website uses analytics, what it records and which cookies it sets. It is the source of truth for that.
Delete things
| What | How |
|---|---|
| One result | Press Delete on the result page, call delete_jobs in MCP, or send DELETE /v1/jobs/:id for a finished job |
| Everything in your account | Delete the account in the dashboard, or send DELETE /v1/account from a signed-in session. Results, keys and any remaining files at Google are removed. |
| A local cache | Delete the folder ~/.scribiz/cache. The CLI caches text, never media. |
| A video you own, from the shared private cache | Send a takedown request. See the DMCA page at /dmca. |
Content you submit
You are responsible for having the right to process what you submit. Scribiz is for personal study, accessibility and your own content. It does not offer downloads of other people's videos. It returns transcripts and subtitles.
Not affiliated
Scribiz is not affiliated with YouTube, TikTok or Instagram.