Log inSubmit server
Video Transcriber logo

Video Transcriber

Transcribes online videos and local files offline with whisper.cpp, supporting 90+ languages and transcript search.

InstallRepositoryWebsite
Full README Compare alternatives Embed widget Claim this listing

Install Video Transcriber

Pick your client, then add it in one click or paste the config.

STDIOPython
~/Library/Application Support/Claude/claude_desktop_config.json
claude_desktop_config.json
{
  "mcpServers": {
    "video-transcriber": {
      "command": "uvx",
      "args": [
        "yt-dlp"
      ]
    }
  }
}

Settings → Developer → Edit Config, merge this into mcpServers, then fully quit and reopen Claude Desktop.

Install not yet sandbox-tested— we haven't run this command through our automated sandbox yet. Not a red flag; we're working through the catalog.

Overview

Video Transcriber MCP server transcribes videos from more than 1,000 platforms or local files with whisper.cpp through a Rust native binary. It can download supported online videos with yt-dlp, process audio with FFmpeg, and run fully offline after the required media and model files are available. Choose among five Whisper model sizes, 90+ languages, and TXT, JSON, or Markdown output. Reach for it when an agent needs local video transcription, privacy-sensitive processing, or searchable transcript content.

Use cases

  • Transcribe videos from supported online platforms
  • Process local meeting or presentation recordings
  • Search across generated transcripts
  • Run privacy-sensitive transcription offline
  • Choose a Whisper model for speed or accuracy

Key features

  • Supports 1,000+ video platforms
  • Transcribes local video files
  • Uses whisper.cpp through a Rust binary
  • Supports 90+ languages and five model sizes
  • Exports TXT, JSON, and Markdown
  • Provides stdio and streamable HTTP transports

Security & health

What it can reach and how reliably it runs — from our own checks, not the maintainer’s claims.

ActiveLast checked 2d ago

Recent health check succeeded.

7/7 checks healthy over the last 45d

Supply-chain signal

8 high-severity advisories on record for this package. Most advisories affect transitive dependencies and may not be exploitable in this server's actual usage — this is a directional signal, not a security audit.

Critical 0High 8Medium 3Low 11

Scanned 12d ago via OSV.dev · yt-dlp (PyPI)

Adoption & maintenance

Factual signals from GitHub, npm, and our automated checks — not a rating.

GitHub stars19Stargazers on the source repository.
Last commit18d agoMost recent push to the default branch.
Directory activity3 viewsConfig copies, upvotes, and views on AllMCPs.

How Video Transcriber works

What Video Transcriber MCP server does

The Video Transcriber MCP server connects an MCP-compatible client to local transcription capabilities built on whisper.cpp and Rust. It accepts videos from supported online platforms, including services such as YouTube, Vimeo, TikTok, and Twitter, as well as local files in formats including MP4, AVI, MOV, and MKV.

The server is intended for agent-driven transcription. A client can request a basic transcription, select a Whisper model, or specify a language. The project description also identifies transcript-wide search as a use case, allowing users to look across generated transcripts rather than processing each file manually.

Five model sizes are available: tiny, base, small, medium, and large. Smaller models favor speed and lower resource use, while larger models generally favor accuracy. Output can be produced as plain text, JSON, or Markdown.

How it works

For an online video, yt-dlp handles downloading from supported platforms. FFmpeg is used for audio processing, and whisper.cpp performs the speech-to-text work locally. This design does not require Python or Node.js at runtime; the packaged application is a native binary.

The Video Transcriber MCP server supports two transports. Stdio is the default for local MCP clients and command-line use. Streamable HTTP exposes an MCP endpoint for remote clients or centralized deployments. The HTTP mode listens on port 8080 by default and can be configured with a host and port.

Transcription is offline once the input media and Whisper model are available. Processing speed depends on hardware, video duration, model size, and audio complexity. The project states that whisper.cpp typically uses less memory and starts faster than Python-based Whisper implementations, but actual performance varies by system.

Setup and configuration

Homebrew installation is available for macOS and Linux and installs the binary with dependencies including CMake, yt-dlp, and FFmpeg. Cargo installation is also supported, but yt-dlp, FFmpeg, and CMake must then be installed separately.

Source builds require Rust 1.85 or newer, yt-dlp, FFmpeg, and clang/libclang. Whisper models can be downloaded with the included script and are stored under ~/.cache/video-transcriber-mcp/models/.

Run video-transcriber-mcp for stdio mode, or use --transport http for HTTP mode. HTTP deployments can set --host and --port; the documented defaults are 127.0.0.1 and 8080. The Claude Code configuration can optionally set RUST_LOG=info.

Tools and capabilities

The Video Transcriber MCP server provides agent-accessible capabilities for:

  • Transcribing videos from 1,000+ online platforms.
  • Transcribing local video files.
  • Selecting one of five Whisper model sizes.
  • Specifying the spoken language, including support for more than 90 languages.
  • Returning TXT, JSON, or Markdown output.
  • Searching across transcripts, according to the project description.
  • Serving MCP over stdio or streamable HTTP.

Limitations and notes

Online transcription depends on yt-dlp support for the target platform and requires network access to obtain the media. Local transcription still requires FFmpeg for audio processing. Larger models need more resources and generally take longer, while long or noisy videos can increase processing time.

The server does not install every dependency in every installation path. Cargo and pre-built binary installations require manual dependency setup, while a source build additionally needs libclang at build time. The Claude Code plugin registers the server and adds a /transcribe skill, but it does not install the binary itself. Search behavior and the exact MCP tool names are not detailed in the provided project material.

Read the full README View source on GitHub

Reviews

No reviews yet — be the first to share how it worked for you.

Frequently Asked Questions about Video Transcriber

Video Transcriber is an MCP server that connects AI clients to whisper.cpp-based transcription for online videos and local files. Its main tools cover video transcription, model and language selection, output generation, and searching across transcripts. It supports more than 1,000 platforms, 90+ languages, and offline processing after media and models are available.

Install it with Homebrew, Cargo, or a pre-built release. Homebrew installs the binary and listed dependencies automatically; Cargo and release binaries require manual installation of dependencies such as yt-dlp and FFmpeg. A source build additionally requires Rust 1.85+ and clang/libclang.

Yes, Video Transcriber works with Claude Code over its default stdio transport. Add the binary to the Claude Code MCP configuration, or install the project’s Claude Code plugin after placing the binary on your PATH. The plugin also adds a `/transcribe` skill.

Yes, it can transcribe local video files offline. Supported examples include MP4, AVI, MOV, and MKV files. FFmpeg is required for audio processing, and a Whisper model must be available locally.

Start the server with `video-transcriber-mcp --transport http`, then connect an MCP client to `http://localhost:8080/mcp`. The default host is `127.0.0.1` and the default port is `8080`; `--host` and `--port` can change them.

Related MCP servers

All in Speech-to-Text All alternatives
Add a badge to your READMEShow it’s listed — live SVG, Markdown or HTML.

AllMCPs Directory Badge

Full Badge Customizer

Showcase your server listing on GitHub or your project documentation. Embed this dynamic SVG badge to highlight official listing status and live engagement.

Badge Style:
Live Dynamic SVG PreviewVideo Transcriber AllMCPs Directory Badge
Markdown (GitHub README)
[![AllMCPs](https://allmcps.com/api/badge/video-transcriber?style=directory)](https://allmcps.com/mcp/video-transcriber)
HTML Embed
<a href="https://allmcps.com/mcp/video-transcriber"><img src="https://allmcps.com/api/badge/video-transcriber?style=directory" alt="Video Transcriber on AllMCPs" /></a>