Tiktok Transcript Scraper

What the tiktok-transcript-scraper MCP server does

The tiktok-transcript-scraper MCP server turns audio from public TikTok videos into structured transcript data. Its speech model listens to the video audio instead of copying TikTok's generated captions, so it can process videos that do not expose captions. The service supports more than 90 languages.

For each video, the response can include cleaned and punctuated text, sentence-level or word-level timing information, speaker labels, and subtitle output in SRT and VTT formats. Segment-level confidence scores help downstream applications assess transcription quality. Optional TikTok metadata can include the author, hashtags, view count, and like count.

A batch may contain failed rows for private, deleted, or region-restricted videos. Those failures are not charged.

How it works

The tiktok-transcript-scraper MCP server is hosted remotely at the Apify MCP endpoint:

https://mcp.apify.com?tools=deapi/tiktok-transcript-scraper

It uses the streamable-HTTP transport. An MCP client connects to that URL, authenticates with Apify, and invokes the available transcription capability for public TikTok videos. The related Apify Actor is available at https://apify.com/deapi/tiktok-transcript-scraper.

Authentication happens before the MCP handshake. Requests without credentials receive 401 Unauthorized, so the service does not provide anonymous access. Clients can authenticate through an Apify OAuth sign-in flow or by sending an Apify API token as a bearer credential.

Setup and configuration

Add the remote server to an MCP client's configuration with its URL and an authorization header:

{
  "mcpServers": {
    "tiktok-transcript": {
      "url": "https://mcp.apify.com?tools=deapi/tiktok-transcript-scraper",
      "headers": {
        "Authorization": "Bearer YOUR_APIFY_TOKEN"
      }
    }
  }
}

Create or retrieve the token in Apify Console under Settings, then API & Integrations. Replace the placeholder with the token used by your client, or choose OAuth if the client supports the browser-based Apify login flow. Keep bearer tokens out of source control and shared configuration files.

Tools and capabilities

The tiktok-transcript-scraper MCP server supports:

• Transcribing audio from public TikTok videos
• Producing punctuated transcript text
• Returning sentence-level or word-level timings
• Identifying speakers in duets and interviews
• Generating SRT and VTT subtitle files
• Providing confidence values for transcript segments
• Including optional author, hashtag, view, and like metadata
• Processing content across more than 90 languages

These outputs make the server suitable for subtitle generation, searchable transcript workflows, and applications that need timing or speaker information from TikTok audio.

Limitations and notes

The input must be a public TikTok video. Private, deleted, and region-locked videos cannot be processed and are returned as failed rows. Authentication is mandatory even though the videos themselves are public.

The service transcribes the audio signal; it does not simply retrieve TikTok's auto-captions. Results may therefore differ from captions shown on TikTok. The README identifies the endpoint as remote and streamable HTTP, but does not document a local deployment mode, additional configuration fields, or named MCP tool identifiers.