Full YouTube videos as clean, readable text
Podcast episodes, tutorials, conference talks, sales webinars — paste the URLs and get the spoken content back as text you can search, quote and export. Up to 30 minutes per video, up to 100 videos per batch.
Long-form is where transcripts pay for themselves fastest. Nobody rewatches a 40-minute interview to find one quote. Everyone will ctrl-F a text file.
Where the text comes from
YouTube's caption coverage is genuinely good, so we use it when it exists: human-authored captions first, automatic captions in the original language second, preferring structured formats that come back clean rather than as overlapping scroll-up fragments.
When a video has no usable track we transcribe the audio. Only the audio stream is pulled for this, never the video — which is why a long video processes faster than you would expect.
Cleanup you do not have to do yourself
Raw caption files are not readable prose. Automatic captions in particular arrive as rolling two-line windows where every line repeats half of the previous one, wrapped in positioning tags. Pasted straight into a document, a ten-minute video becomes four thousand words of duplicated fragments.
We de-duplicate the overlap, strip the markup and return continuous punctuated text. That is the difference between a transcript you can hand to a colleague and a transcript you have to spend twenty minutes cleaning first.
What long-form transcripts get used for
- Finding the quote. Search the text, get the sentence, cite it correctly instead of paraphrasing from memory.
- Show notes and summaries. The transcript is the raw material; the summary writes itself once the words exist.
- Repurposing into short-form. A 30-minute episode usually contains four or five self-contained segments. See repurposing.
- Competitive and market research. Sales webinars and product walkthroughs state pricing, positioning and objection handling out loud.
- Accessibility and archives for content you own — making your own back catalogue searchable is the least glamorous and most durable use.
Handling videos over the 30-minute limit
The ceiling is per item, not per batch, so a long recording is a splitting problem rather than a dead end. If you control the source — your own webinar, your own podcast — upload or link the segments separately and paste them as one batch; they process in parallel and you stitch the text back together in order.
For a long video you do not control, the pragmatic move is to decide whether you need all of it. Most two-hour livestreams contain twenty minutes worth reading, and the chapter markers usually tell you which twenty.
Cost on long videos
Per-second billing cuts both ways. On short-form it saves you the rounding; on long-form it means a 28-minute video really does cost 28 minutes of balance. The $9 / 300-minute pack is about ten half-hour videos, or several hundred Shorts. Plan the pack around which end of that range you actually live at.
The hard ceiling is 30 minutes per item, checked before anything is charged. A two-hour livestream will be rejected rather than half-processed.
Try it on your own links
Ten free minutes, no card. Paste a batch of TikTok or YouTube links and read the scripts.
Start extracting →Questions people ask
How long can a YouTube video be?
30 minutes per video. The check happens before billing, so an over-length video costs you nothing.
Do you use YouTube's own captions or transcribe the audio?
Captions when a usable track exists — human-authored preferred, automatic captions otherwise — and speech recognition on the audio when there is nothing to use. Either way the output is cleaned into continuous prose.
Can I transcribe a whole playlist or channel?
Not as a single link. Paste the individual video URLs, up to 100 per batch. For TikTok, whole-profile fetching is supported — see the profile guide.
Are timestamps included?
The output is continuous readable text rather than a timestamped subtitle file. Duration, author and engagement figures come attached to each item, and the batch exports to CSV.