TikTok to Text
Extract audio and transcribe speech from a public TikTok link or a local media file.
Whisper tiny runs on this device. First use downloads about 41 MB of speech models plus 22 MB of runtime from Hugging Face / this site, cached when supported. Generative summaries download an additional ~284 MB English model and can use significant memory. Processing may take several minutes, especially on phones. Audio and text are not sent to an AI service. TikTok extraction uses our download proxy (100 MB cap); your own files stay local. Review recognition and summaries for accuracy.
Processing your video...
Transcription
Your transcription will appear here...
Summary
AI summary will appear here...
Recommended Tools
Using TikTok to Text
Which inputs does TikTok to Text support?
Choose Audio only, Transcription, or Transcription + Summary. Paste a TikTok URL or select an audio/video file up to 100 MB; local AI modes require confirming the model download.
What is an example of the result?
Load the sample English speech and choose transcription to receive a copyable TXT transcript. A video with an audio track can first produce an M4A audio download; a silent video cannot provide speech.
Which limits or failure conditions matter?
Local transcription accepts up to 30 minutes of decoded audio. Model downloads can be large and first use can be slow. Whisper output and English AI summaries need proofreading; short summaries preserve the source text, and completed transcripts survive summary failure.
Where is this task processed?
A submitted TikTok URL goes to the server to retrieve media. Local-file audio extraction, Whisper transcription, and summarization run in this browser after engines and models download; local media is not sent to a paid transcription API.