Skip to main content
Glama
nnnoel

ytranscript-mcp

by nnnoel
OverviewSchemaSourceRelated ServersDiscussionsAdmin
TypeScript
Remote

ytranscript-mcp

An MCP server that gives Claude, Cursor, VS Code, and other AI tools the transcript of any YouTube video.

  • Works from anywhere. Runs fine on AWS, Lambda, Vercel, n8n Cloud, and other cloud hosts. There are no proxies to buy and no IP bans to work around. YouTube blocks most cloud IPs for transcript requests; the yTranscript API handles that for you.

  • Handles videos without captions. If a video has no captions, it's transcribed with AI speech-to-text.

  • Returns the language actually spoken by default, including on auto-dubbed videos. Pass lang to pick a specific caption track.

  • Keeps long videos manageable. Choose plain text or timestamped lines, and cap the length so a two-hour video doesn't flood the model's context.

Setup

  1. Get a free API key at ytranscript.com/developers. The free plan includes 50 units a month, no card required.

  2. Add the server to your client using one of the configs below.

Claude Code

claude mcp add ytranscript -e YTRANSCRIPT_API_KEY=yk_live_your_key -- npx -y ytranscript-mcp

Claude Desktop

Add this to claude_desktop_config.json. Open it from Settings → Developer → Edit Config.

{
  "mcpServers": {
    "ytranscript": {
      "command": "npx",
      "args": ["-y", "ytranscript-mcp"],
      "env": { "YTRANSCRIPT_API_KEY": "yk_live_your_key" }
    }
  }
}

Cursor

Add this to ~/.cursor/mcp.json, or to .cursor/mcp.json in a project:

{
  "mcpServers": {
    "ytranscript": {
      "command": "npx",
      "args": ["-y", "ytranscript-mcp"],
      "env": { "YTRANSCRIPT_API_KEY": "yk_live_your_key" }
    }
  }
}

VS Code

Add this to .vscode/mcp.json:

{
  "servers": {
    "ytranscript": {
      "command": "npx",
      "args": ["-y", "ytranscript-mcp"],
      "env": { "YTRANSCRIPT_API_KEY": "yk_live_your_key" }
    }
  }
}

Any other MCP client that runs stdio servers works the same way: run npx -y ytranscript-mcp with YTRANSCRIPT_API_KEY set.

Related MCP server: YouTube Transcript MCP Server

Tool: get_transcript

Argument

Required

Description

video

yes

A YouTube URL (watch, youtu.be, shorts, embed) or an 11-character video ID

lang

no

Language code of the caption track to fetch, such as en or es. Omit it to get the video's original language

format

no

text (default) returns plain text. timestamped returns one line per segment with [m:ss] times

max_characters

no

Cut the transcript at about this many characters. The response says how much was left out

Things to ask your assistant:

  • "Summarize this video: https://youtu.be/..."

  • "What does the speaker say about pricing, and at what timestamps?"

  • "Pull the transcript of this Spanish video and translate it to English."

Every response ends with the units used and your monthly total, so you can keep an eye on your quota.

Pricing

A transcript from captions costs 1 unit. A video without captions, transcribed with AI, costs 15 units. Failed requests are free.

Plan

Price

Units per month

Free

$0

50

Starter

$9/mo

1,000

Developer

$29/mo

10,000

Scale

$79/mo

50,000

See ytranscript.com/developers for details.

Related

  • ytranscript-api is the JavaScript/TypeScript client this server is built on.

License

MIT

Available Tools

1 tool
get_transcriptGet YouTube transcriptA
Read-only

Get the transcript of a YouTube video. Accepts a video URL (watch, youtu.be, shorts, embed) or an 11-character video ID. Returns the video's spoken language by default. Videos without captions are transcribed with AI speech-to-text. Use format=timestamped when you need to cite or jump to specific moments.

ParametersJSON Schema
NameRequiredDescriptionDefault
langNoLanguage code of the caption track to fetch, e.g. 'en' or 'es'. Omit to get the video's original language.
videoYesYouTube video URL or 11-character video ID
formatNo'text' (default): one block of plain text. 'timestamped': one line per segment with [m:ss] times.
max_charactersNoTruncate the transcript to about this many characters. Useful for very long videos.

TDQS

A4.4/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations only declare readOnlyHint and openWorldHint, so the description adds meaningful behavioral context: it returns the video's spoken language by default and falls back to AI speech-to-text when captions are unavailable. This goes beyond the structured annotations and helps set expectations about the transcript source.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Four short sentences with no redundant content. The core purpose is front-loaded, and each sentence adds distinct value: input formats, output default, fallback behavior, and parameter use case. Nothing could be removed without losing useful information.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the tool's simplicity, four parameters, read-only annotations, and full schema coverage, the description covers the essential context an agent needs to call the tool correctly. It could mention edge cases like invalid IDs or how errors are surfaced, but these are minor gaps for a read-only transcript fetch tool.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so the baseline is 3. The description adds extra nuance not fully captured by the schema: it clarifies accepted URL variants for the video parameter and explains the default language behavior. The format guidance ('timestamped when you need to cite or jump to specific moments') also adds practical parameter semantics.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states a specific verb and resource: getting the transcript of a YouTube video. It enumerates accepted input formats (watch, youtu.be, shorts, embed URLs or 11-character IDs) and states the default behavior (spoken language), making the tool's purpose unambiguous even without sibling tools.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description gives clear context for when to use the tool: whenever a YouTube transcript is needed. It also provides format-level guidance ('Use format=timestamped when you need to cite or jump to specific moments'), though it does not explicitly discuss alternatives or when not to use the tool, which is less necessary given there are no siblings.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

  1. 1 tool updatev0.1.0
    • First observedget_transcript
Deploy Server
One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

TDQS

A4.4/5.0

Scored across 1 tool

Disambiguation5/5

With only one tool, there is no possibility of overlap or confusion. The tool's purpose is clear and singular.

Naming Consistency5/5

The single tool name follows a clear verb_noun pattern (get_transcript), consistent with common MCP conventions. No inconsistency can exist with one tool.

Tool Count3/5

One tool is borderline thin for an MCP server, but for a narrowly scoped YouTube transcript service it is not unreasonable. It would benefit from additional tools like listing available languages or transcript metadata.

Completeness4/5

The tool covers the primary use case well: accepts various URL formats, detects language, falls back to AI speech-to-text, and supports timestamped output. Minor gaps exist such as no explicit language selection or list of available transcripts, but these are not critical for the core purpose.

Resources

Related MCP Connectors

Related MCP Servers