Evil0ctal/Douyin_TikTok_Download_API

Project overview

Its documented workflow combines scheduled recollection, persistent PostgreSQL archiving, bulk media downloads, and a self-healing guest-identity pool for ongoing Douyin and TikTok collection.

Project type

MCP · AI Search · Data Processing

Use cases

Automation

Deployment

Refer to project documentation

License

Apache-2.0

Best for

• Developers, AI engineers, or operations teams that need self-hosted Douyin and TikTok collection, local media downloads, persistent PostgreSQL archiving, and access through web, API, MCP, or CLI interfaces.
• Teams prepared to operate a service that depends on Douyin or TikTok upstream access and may need to respond to expired cookies, changed signature algorithms, and rate limits.

Key capabilities

• Fetches posts, author profiles, author posts, liked posts, playlists, comments, and comment replies from both platforms; follower and following lists are additionally available for TikTok.
• Downloads video and image albums using clean streams published by the source platforms rather than stripping watermarks.
• Archives parsed content in PostgreSQL so collected posts can remain available after deletion from the upstream platform.
• Watches an author or post and recollects it on a timer.
• Uses a headless browser to mint guest identities and replenishes the pool when usable identities run low.
• Exposes the same data service through a web console, REST API, MCP endpoint, and command-line interface.
• Accepts short links, post URLs, shared caption text, and bare post IDs; follows short links, extracts embedded links, and validates IDs before making an upstream request.
• Downloads media to the user's disk in bulk by author, skips existing items, and performs duplicate cleanup.

Limitations and risks

• Douyin follower and following endpoints are not registered because those lists require a signed-in session; TikTok has the additionally documented follower and following support.
• The public demo is read-only, limited to 30 requests per 10 seconds, and is not a service on which users should build.
• The v5 rewrite does not support Bilibili.
• Expired cookies, changed signature algorithms, and endpoint rate limits can disrupt collection behavior, so operation depends on continued upstream compatibility and access.
• Database migrations have no automatic downgrade. Backing up the database before upgrading is necessary for the documented reliable rollback procedure.
• Users are responsible for complying with platform terms and applicable law, respecting collected-content subjects, and avoiding unauthorized redistribution.

Getting started

• The documented first-success sequence is to download install.sh, inspect install.sh, and run bash install.sh. The guided script asks a few questions, detects the distribution, checks for Docker, scales resource limits, and brings up the stack.

Evidence and sources

• GitHub project description: 🚀 Self-hosted TikTok & Douyin scraper and no-watermark video downloader — async REST API, MCP server, CLI and web console for posts, profiles, comments and playlists. Self-healing…
• README: Open source, free, and it runs on your own machine — no signup, no quota, nobody else in the path. It fetches posts, authors, comments and search, downloads video and image albums…
• README: v4's real problem was never a shortage of features — it was that **the API would die quietly and nobody would know**. A cookie expires, a signature algorithm changes, an endpoint…
• README: | **Docker Compose** (recommended) | Almost everyone, production included | [Installation · First install](./documents/en/02-installation.md#first-install) | | By hand | Docker is…
• Release: v5.1.0

Source updated at: 2026-09-15T18:05:08Z

Open original GitHub project