Paste a TikTok link and your assistant gets the spoken words as text, whether or not the video ever had captions.
get_transcriptsearch_videoslist_channel_videoslist_playlist_videosget_creditsAny MCP client connects the same way: point it at the server URL and send a TranscriptFetch API key as a bearer header (on claude.ai, OAuth replaces the key).
https://transcriptfetch.com/mcpnpx mcp-remote https://transcriptfetch.com/mcp \
--header "Authorization: Bearer tf_live_YOUR_KEY"The server exposes the same registry to every client; these are the calls that matter for TikTok, described the way your assistant sees them.
get_transcript(video)Takes a full tiktok.com/@creator/video/... link or the short vm.tiktok.com and vt.tiktok.com links the app's share sheet produces, so a link sent from a phone works as-is. Because TikTok rarely has a caption file, the fetch normally runs as an AI Fallback Transcription job, and on a 30-to-60-second clip that finishes fast enough to feel synchronous. The result comes back as timestamped segments with the detected language: the words as spoken, untranslated.
search_videos(query, platform: "tiktok")Keyword search on TikTok: the matching posts with their captions, play counts and URLs, each one ready for get_transcript. The way in when the user has a topic rather than a link.
list_channel_videos(channel)Takes a profile URL (tiktok.com/@creator) and returns the creator's recent posts, up to 50 per call, so an agent can read what someone has been saying lately without anyone collecting links by hand.
list_playlist_videos(playlist)Takes a TikTok playlist URL (tiktok.com/@creator/playlist/name-id) and returns the posts in it: the series a creator grouped themselves, in their order.
get_credits()The free balance check. Worth calling before pointing an agent at a folder of 80 saved links: TikToks are cheap at one credit each, but an unattended batch still deserves a budget.
TikTok's auto-captions are opt-in per upload, so the majority of videos expose no caption file. A tool that only extracts existing captions comes back empty on most TikTok links, and that is not a bug, there is genuinely nothing to extract. This server transcribes the audio instead, so the caption file's absence does not matter.
Usually not as files: TikTok's captions are opt-in and most creators skip them or burn text into the frames. The server transcribes the audio with AI, so a missing caption file does not produce a missing transcript.
Yes. search_videos accepts TikTok as a platform, and the channel and playlist tools accept TikTok profile and playlist URLs. Every returned URL can be sent straight to get_transcript.
No speech means no transcript, and the fetch returns empty rather than inventing text. Failed and empty fetches are never billed.
Yes. vm.tiktok.com and vt.tiktok.com short links resolve to the underlying video before fetching, so you can paste straight from the share sheet without opening the link in a browser first.
The transcription detects the spoken language on its own and returns the words as said; nothing is translated. The detected language is reported alongside the timestamped segments.
50 free credits a month. Caption fetches cost one credit; AI Fallback Transcription is billed by audio duration, only on delivery. Longer TikToks use more AI Fallback Transcription credits than shorter ones.
Sign up, grab a key, paste one config block into your client. Most TikToks transcribe in seconds.