AI Video Platform · Engine test
Speech recognition test
Test the first AI component by giving it a short video or audio file. The browser reads the audio, then Whisper creates a transcript with timestamps.
Use a direct media link
Paste a direct HTTPS link to an MP4, WebM, MP3 or WAV file. This is not the normal YouTube watch-page URL.
YouTube watch links cannot be processed by this test page. The media host must allow browser access (CORS); some websites block this. The page does not bypass those restrictions.
For this first test: maximum 100 MB and 10 minutes. Model files are downloaded once from Hugging Face; transcription then runs on your device.
Ready when you are.