TikTok Transcript & Instagram Reels Transcript
Pricing
from $1.50 / 1,000 transcript from captions
TikTok and Instagram Reels transcripts (plus YouTube Shorts) as text, timestamped segments or SRT. No login or API key. Platform captions first, AI speech-to-text only when there are none.
TikTok, Instagram Reel/post, or YouTube Shorts URLs. Mix platforms freely. One dataset item is produced per URL; a failure on one URL never stops the others.
Leave as "auto" to get each video's original language (recommended: fastest and most reliable). Set an ISO-639-1 code (e.g. "en", "es", "pt") to prefer that caption language when the platform offers it; YouTube can machine-translate its captions into it. Falls back to the original language when the requested one is unavailable.
If a video has no platform captions, download its audio and transcribe it with AI speech-to-text. Billed per started minute of video (transcript-ai-minute event). Videos with captions are never sent to AI.
"auto" uses a local faster-whisper model inside the Actor (no external API). OpenAI gpt-4o-mini-transcribe is used only when the Actor owner has configured it and you are on a paid Apify plan.
Local faster-whisper model for the AI path. "base" (default) is noticeably more accurate; "tiny" is a little faster. Same price either way.
Which representation is copied into the top-level "output" field for convenience. All three (text, segments, srt) are always included in every dataset item regardless of this setting.
Used only when a platform blocks direct requests (YouTube always does; TikTok and Instagram sometimes). Defaults to Apify's residential proxy. Set to no proxy to force direct requests only.
{ "useApifyProxy": true, "apifyProxyGroups": [ "RESIDENTIAL" ]}