Who operates this site
textfromvideos.com is operated by the operator of this website. A separate company name and mailing address are not published here.
What the service does
The site lets you paste a public TikTok or Instagram Reels URL into a form and receive a written transcript. It does not accept uploaded video files. You do not create an account, and the site does not take payments. If a public video already has official or automatic captions, those captions can be turned into text. If there is no usable caption track, the audio is transcribed with Deepgram. You can copy the result or download TXT, SRT, or VTT. Videos longer than three minutes are rejected. You can transcribe up to three new public videos per IP address each UTC day. A cached result from the last seven days does not count toward that limit.
Information we collect
When you request a transcript, we collect the video URL you submit, including the canonical form we resolve it to, and your IP address. The IP address is used to enforce the free daily limit and a cap on repeated failed attempts. If a transcript is produced, we store the plain text and the timestamped subtitle file, along with the platform, video id, language, and whether the text came from captions or speech-to-text.
We also record first-party analytics events named transcribe_start, transcribe_success, and copy_click, together with your IP address and a small event payload. Quota and failure events needed to run the limits, including transcribe_quota and transcribe_failure, are stored in the same way. Job records keep the submitted URL and the status of that request.
We do not ask for your name, email address, or a password. The site does not create user accounts and does not process payments.
How transcripts are stored
A successful transcript is stored so the same public video can be returned from cache for seven days. During that window, pasting the same link again returns the earlier text, SRT, and VTT without a new speech-to-text run, and without consuming the daily limit. After that seven-day window, a later request is treated as a new transcription. The site does not currently publish an automatic deletion schedule for older transcript rows, job records, or events, and it does not offer a self-service history or deletion form, because there are no accounts.
Speech-to-text and other processors
To transcribe a public link, our servers fetch the video from TikTok or Instagram. Those platforms receive a request from our servers when you submit their URL. If captions are not available, the audio is downloaded to a temporary file, sent to Deepgram for speech-to-text, and then removed from the temporary directory. Deepgram processes that audio under its own terms. Only the resulting transcript text and subtitle file are kept for the cache described above. The video file itself is not stored as a product feature.
Advertising, cookies, and third-party analytics
This site uses Google AdSense to display advertisements. Google and other third-party vendors may use cookies, including the DoubleClick cookie, to serve personalized ads based on your prior visits to this website or other websites on the internet. You can opt out of personalized advertising at any time through Google Ads Settings, and you can review how Google uses data in advertising on the Google Advertising Privacy & Terms page.
The transcript tool itself does not require account cookies. The first-party analytics events listed above are stored on our servers. The site can also optionally load Google Analytics 4 or Plausible if those services have been configured. When those scripts are present, the relevant provider may set cookies or collect usage data under its own policy.
How we use this information
We use the submitted URL and the resulting transcript to provide the tool, to serve cache hits for seven days, and to generate TXT, SRT, and VTT downloads. We use IP addresses to enforce the free daily limit and failure caps, and to record first-party analytics. We do not sell this information, and we do not use the transcript tool to run accounts or payments.
Your choices
If you do not want a video transcribed, do not submit the URL. A cache hit only returns an existing transcript for seven days; it does not start a new speech-to-text run. Because there are no accounts, there is no login where you can view or delete a personal history. For privacy questions, you can email yangkai422@gmail.com.
Children
This is a general-purpose website. It is not directed at children, and we do not knowingly collect personal information from children.
Changes
If this policy changes, we will update this page. The date at the top reflects the current version. The terms and about pages describe how the tool works in more general terms.