This is a short NASA video about asteroid 2024 YR4, run through our tool on 2 October 2026. The talking runs to 0:35, and this is the whole result, unedited.
One honest note: it is an Instagram Reel, not a TikTok. We don't copy other people's TikToks onto our site, and NASA, whose videos are public domain, posts short videos on Instagram. Both kinds of link go through the same step in our tool: speech-to-text listens to the audio and writes down what is said. So a TikTok you paste comes back in exactly this form.
What happens when we detect an asteroid that could pose a threat to Earth? — NASA on Instagram. Watch itWords and times come from the video's audio, by speech-to-text — the same step a TikTok link goes through.
0:00This asteroid 2024 YR4 might get interesting.
0:03There was a chance that it could impact Earth.
0:07The Atlas Survey, operated by the University of Hawaii, finds about 200 to 250
0:12near-Earth objects per year.
0:14This object was sort of a garden variety new NEO.
0:17There was nothing about it when we first saw it that suggested that it would be especially dangerous.
0:21The impact probability climbed.
0:25It started to increase with more data.
0:27That's when things really kind of heated up for us.
0:29On December 22, 2032, there was a chance that it
0:34could impact Earth.
Reading it like a creator
The hook is the first line. “This asteroid 2024 YR4 might get interesting.” Seeing it written down makes it easy to study how the video opens.
Talk stays talk. “Sort of a garden variety new NEO” and “things really kind of heated up” are kept as spoken. Fillers and loose phrases are what make a short video sound natural, and the transcript doesn't tidy them away.
Numbers are digits. “200 to 250” and “December 22, 2032” came out the way you would type them.
Lines break at pauses. The sentence at 0:29 runs into 0:34 because the speaker paused in the middle of it.
No speaker names. Speech-to-text writes down words, not who said them. If more than one person talks, add names yourself before you quote it.
What to check before you use it
Names and special terms. The survey at 0:07 is officially called ATLAS, in capitals. Speech-to-text heard the word and wrote “Atlas”. Names like this are worth checking against the video.
The first and last lines. If a video starts or ends with music, speech-to-text sometimes adds a stray word like “you”. Delete it.
Music-only TikToks. If no one talks, there is nothing to write down. The tool then tells you it couldn't hear anyone talking, and that try does not count. See the Reels example for what speech-to-text writes on music.
A transcript vs TikTok's captions
TikTok can show captions on a video so people can watch with the sound off. They appear a few words at a time, on screen. A transcript gives you all the words at once, as text you can copy, search, quote or save as a file. That is what you need to reuse a video's script.
What you could do with it
Study hooks. Collect the first lines of videos in your niche and compare them. More on extracting hooks.
Write a caption. Start from the spoken words instead of a blank box.
Lines of spoken text, each with the time it starts in the video, plus a short summary. You can copy it or download it as TXT, SRT or VTT. The example on this page is real output.
Why is the example not a TikTok?
We don't copy other people's TikToks onto our site, and NASA, whose videos are public domain, posts short videos on Instagram. TikTok and Instagram links go through the same speech-to-text step in our tool, so a TikTok comes back in exactly this form.
Does it pick up text written on the screen?
No. The transcript comes from the audio. Text that only appears on screen, and the music, are not in it. A TikTok with no talking gives you little or nothing.
Is it free?
Yes, to try. Without an account you get 1 free video every 30 days, and signing in free gives you 1 more. Free results show the first minute; TikToks of a minute or less come back in full. Pro ($13.50/month) gives full-length transcripts and up to 10 links at once.