Creativity · MCP — server

MCP TikTok Server

TikTok's Content Posting and Business APIs let approved apps upload videos, fetch video insights, and manage creator workflows. The MCP TikTok Server wraps those endpoints so Claude can draft captions, schedule uploads, and pull post-performance stats for creators and brands.

MCP facts

Kind
server
Ecosystem
anthropic-mcp
Language
Python
Transports
stdio

Capabilities

  • Tools: list_videos, get_video_insights, upload_video, query_creator_info
  • OAuth2 with user-granted scopes
  • Supports TikTok Ads and TikTok for Business APIs

Install

pip install mcp-tiktok

Configuration

{
  "mcpServers": {
    "tiktok": {
      "command": "uvx",
      "args": ["mcp-tiktok"],
      "env": {
        "TIKTOK_CLIENT_KEY": "<client-key>",
        "TIKTOK_CLIENT_SECRET": "<client-secret>",
        "TIKTOK_ACCESS_TOKEN": "<user-access-token>"
      }
    }
  }
}

Frequently asked questions

Can the model publish videos?

Yes — with the video.publish scope. Treat this as a high-risk action and require explicit approval at the MCP client layer.

Is the app subject to review?

Yes, TikTok reviews apps requesting the posting and business scopes. Your MCP server will need production-ready compliance.

Does it work with TikTok Ads Manager?

A separate set of tools under TikTok for Business lets the model pull ad insights; policies differ from organic content posting.

Sources

  1. TikTok Content Posting API — accessed 2026-04-20
  2. Model Context Protocol — accessed 2026-04-20