Introducing Transcript Segmentation - Up to 4 Hours for Facebook, 2 Hours for TikTok
You can now transcribe long videos in 15-minute chunks. Select any segment up to 4 hours (Facebook) or 2 hours (TikTok) and get your transcript in minutes.
Long Videos Now Fully Supported
Until now, our AI transcription worked best on short videos reels, clips, and highlights under 15 minutes. Today we're introducing Transcript Segmentation, which lets you transcribe any long-form Facebook or TikTok video by selecting the exact 15-minute window you need.
Why We Built It
Facebook Live streams, long-form interviews, lectures, webinars, and watch-party recordings can run anywhere
from 30 minutes to several hours. Downloading and transcribing that entire audio file puts enormous strain on
our GPU workers, takes longer for you to wait, and costs more to run.
Segmentation solves this cleanly: instead of processing the whole file, our worker downloads only the 15-minute window you asked for — using ffmpeg to cut precisely — then runs Whisper on that small chunk. The result is faster, cheaper, and more reliable.
How It Works?
Limits by platform
- These limits reflect typical video lengths on each platform. Facebook supports long live streams and
- watch parties; TikTok's longer format tops out at around 60–90 minutes in practice.