← Back to directory
B

Bilibili MCP Server

Community
A local MCP server for AI agents to read Bilibili subtitles and comments
Category
Other #42 of 230
Stars
★ 34 Popular
Transport
stdio (local process)
Runtime
Node.js 20+ · Python 3.9+
Credentials
API key / credential required
License
Apache-2.0
Last commit
Tools
9
64FMRS · C

A feature-rich local MCP server for Bilibili content. It requires credentials, supports search, subtitles, comments, favorites, and optional local ASR.

Strongest · Documentation 16/20 Weakest · Reliability 10/20

Reliability
10/20
Security and permissions
14/20
Maintenance
11/20
Documentation
16/20
Setup experience
13/20
Why each score
Reliability 10/20
README claims ten tools, a verified acceptance chain, Vitest tests and npm test, but no CI workflow or committed test content was supplied for static verification, so the calibration cap of 12 applies; error codes (ASR_NOT_READY etc.) and limits are well disclosed, deducted for lack of verifiable runtime/handshake evidence.
Security and permissions 14/20
Good stated practices: hidden-prompt local credential entry, doctor never echoes cookies, subtitle-domain and HTTPS Bilibili-CDN allowlists, no access-control bypass, UGC treated as data; however the actual code is unavailable, cookies are stored plaintext in config. with no OS-level encryption by the project's own admission, and network boundaries cannot be statically confirmed. No red-line findings, so not scored 0-4.
Maintenance 11/20
Version 1.13.1, CHANGELOG, latest-release link, Issues and Discussions channels exist, but the project is small (12 stars); GitHub license metadata is NOASSERTION (repo says GPLv3, an inconsistency) and there is no evidence of a security-response channel or dependency-update policy — active with governance gaps.
Documentation 16/20
Strongest dimension: layered docs (CN/EN README, client setup, tool reference, changelog), install, credential acquisition, worked examples, error codes, limits, and privacy statements; deducted because all evidence is declarative, the referenced parameter tables were not reviewable, and some assumptions (Python only needed for ASR) are buried.
Setup experience 13/20
Node 20+ prerequisite, one-line npx registration, an agent-assisted install prompt, and a setup/check/doctor verification flow are clear and include a client-config guide; deducted for the multi-step interactive cookie configuration, no verified client config examples or screenshots, and the static cap of 15.

Static review · not runListed 2026-09-08

Read the FMRS scoring method →

Fit and risk

What it can accessReads local filesUses the network

Best for

  • Creators or researchers who want AI assistants to extract text content from Bilibili videos
  • Users who need to batch organize Bilibili favorites or subtitle data

Not for

  • Users not willing to provide Bilibili Cookie credentials
  • Anonymous use cases without accessing private content such as favorites
  • Users requiring long transcriptions that exceed resource limits

Required permissions

  • Requires Bilibili login credentials (Cookie) for features like search and favorites
  • Credentials stored locally in config file (~/.bilibili-mcp/config.json)
  • Requests are sent only to official Bilibili endpoints

Risks and side effects

  • Credentials stored locally without OS-level encryption
  • High-frequency calls may trigger Bilibili risk control or rate limiting
  • This is a third-party tool, not an official Bilibili service

Setup

Before you start

Runtime:Node.js 20+ · Python 3.9+

  1. Ensure Node.js 20+ is installed.
  2. Add the stdio server to client config: command npx, args ['-y','@xzxzzx/bilibili-mcp@latest'].
  3. Run npx -y @xzxzzx/bilibili-mcp@latest setup to configure credentials (Cookie), then run check to confirm.
  4. Restart/reconnect the client, call check_bilibili_credentials to verify login, then test with search_bilibili_videos.
claude_desktop_config.json
{
  "type": "stdio",
  "command": "npx",
  "args": [
    "-y",
    "@xzxzzx/bilibili-mcp@latest"
  ]
}

Shown for Claude Desktop. Other clients may use a different file or key (VS Code uses "servers") — the configurator below converts it.

Check that it works

After reconnecting the client, call check_bilibili_credentials and confirm configured: true and logged_in: true; then call search_bilibili_videos once with any topic — returning a video list confirms the setup works.

Troubleshooting

  1. If login fails, check credential configuration (run `setup` to force reconfigure)
  2. If ASR unavailable, ensure the model is installed via `setup` and `doctor --json` reports `asr.status: ready`
  3. If the MCP server is unavailable, check client configuration and network connectivity

Things to try

Once connected, you can ask your AI assistant things like:

  • Read the subtitles of BV1Eb411u7Fw with timestamps, then fetch the video's top comments and replies.
  • Search Bilibili for videos about 'discrete mathematics' and list 5 candidates sorted by platform relevance, with title, uploader, duration, and BVID.
  • Read the P4 subtitles of BV1Eb411u7Fw and search for the keyword '函数', returning match context, timestamps, and direct Bilibili moment links.
  • Iterate all favorite folders of my logged-in account, following next_cursor until finished, and report skipped_count.

Tools 9

search_bilibili_videos read-only
Search Bilibili videos by topic and get a list of candidates
list_bilibili_favorite_videos read-only
Iterate through the current account's favorites and read videos
get_video_transcript read-only
Retrieve full transcript, keyword search, or trigger local ASR fallback
get_video_metadata read-only
Get structured metadata like title, author, view count
get_video_comments read-only
Get comments and replies for a video
get_video_chapters read-only
Get video chapters and progress bar segments
get_credential_setup_instructions read-only
Guide user through configuring Cookie
check_bilibili_credentials read-only
Check if Cookie is configured and logged in
Show 1 more tools
check_mcp_update read-only
Check if the MCP package has an update

Use cases

Read subtitles and comments of Bilibili videos, supporting timestamps and keyword search
Search videos by topic and get metadata
Iterate your own favorites for batch processing
Transcribe videos without subtitles using local ASR (requires model installation)

Supported clients

Claude Desktop

Listed from the project's documentation, not tested by this site.

Overview

Bilibili MCP is a local MCP server that enables AI agents to read Bilibili content: fetching subtitles and comments, searching videos, browsing favorites, and transcribing videos without subtitles via an optional local ASR model.

Similar servers

BioMCP 75 · B

One binary. One grammar. Evidence from the biomedical sources you already trust.

★ 659 · Tools 11 Compare with this →

Source revision 5ad5b7459fcd Data synced 2026-10-19 Read the FMRS scoring method