Skip to main content
Glama

youtube_audio_transcript

Speech-to-text for YouTube audio. Use it when a video has no captions — or when you want a transcript of what was actually spoken rather than YouTube's published captions. Priced per started minute of audio. Costs 2 credits/min of audio. Empty results and failures are never charged. Pass cache=true for a free 24h cache hit (default always fresh).

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "url": "<url>"
    },
    "name": "youtube_audio_transcript"
  }
}

Input Schema

NameRequiredDescriptionDefault
urlYesPublic YouTube video URL, e.g. https://youtube.com/watch?v=ID. Not a TikTok/Instagram/Facebook URL. The URL platform must match this endpoint's platform. Do not pass cross-platform URLs, e.g. YouTube to TikTok, Instagram to Facebook, LinkedIn to X/Twitter, or Pinterest to Rumble.
cacheNoSet true to serve from the 24h response cache (0 credits on hit). Default false — always fetch fresh.
fieldsNoOptional comma-separated keys to keep on each result row (or on the object for single-item tools), e.g. id,url,publishedAt,engagement. Pagination keys are always kept. Does not change the price.
compactNoDefault true here: drop signed media URLs (video, image, avatar links and their expiry stamps) to keep the response small. Ids, permalinks, dates, text and metrics stay. On Threads, Facebook, Meta Ad Library and TikTok Shop tools request timings are dropped too. Pass false for the full body.
languageNoPreferred caption language as an ISO code, e.g. "en". Defaults to auto-detect.
maxCreditsNoRefuse before STT when estimatedCredits would exceed this (400 cost_exceeds_max, 0 credits). The estimate is on every success and, when the extract miss is retryable, as estimatedCreditsIfRetried.

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed1 schema field changed
    • changedInput schema / properties / compact / description
      Previous value: -"Default true here: drop signed media URLs (video, image, avatar links and their expiry stamps) to keep the response small. Ids, permalinks, dates, text and metrics stay. Pass false for the full body."New value: +"Default true here: drop signed media URLs (video, image, avatar links and their expiry stamps) to keep the response small. Ids, permalinks, dates, text and metrics stay. On Threads, Facebook, Meta Ad Library and TikTok Shop tools request timings are dropped too. Pass false for the full body."
  2. Changed2 schema fields changed
    • addedInput schema / properties / compact
      Added value: +{
      +  "default": true,
      +  "description": "Default true here: drop signed media URLs (video, image, avatar links and their expiry stamps) to keep the response small. Ids, permalinks, dates, text and metrics stay. Pass false for the full body.",
      +  "type": "boolean"
      +}
    • addedInput schema / properties / fields
      Added value: +{
      +  "description": "Optional comma-separated keys to keep on each result row (or on the object for single-item tools), e.g. id,url,publishedAt,engagement. Pagination keys are always kept. Does not change the price.",
      +  "type": "string"
      +}
  3. Changed1 schema field changed
    • removedInput schema / properties / url / format
      Removed value: -"uri"
  4. Changed1 schema field changed
    • changedInput schema / properties / cache / description
      Previous value: -"Set true to serve from the 24h response cache (0 credits on hit). Default false — always fetch fresh. Envelope includes cached + cachedAt on hits."New value: +"Set true to serve from the 24h response cache (0 credits on hit). Default false — always fetch fresh."
  5. Changed1 schema field changed
    • changedInput schema / properties / maxCredits / description
      Previous value: -"Refuse before STT when estimatedCredits would exceed this (400 cost_exceeds_max, 0 credits)."New value: +"Refuse before STT when estimatedCredits would exceed this (400 cost_exceeds_max, 0 credits). The estimate is on every success and, when the extract miss is retryable, as estimatedCreditsIfRetried."
  6. Changed3 schema fields changed
    • changedInput schema / properties / cache / description
      Previous value: -"Set true to serve from the 24h response cache. Default false — always fetch fresh data."New value: +"Set true to serve from the 24h response cache (0 credits on hit). Default false — always fetch fresh. Envelope includes cached + cachedAt on hits."
    • changedInput schema / properties / maxCredits / description
      Previous value: -"Refuse before STT when estimated credits would exceed this."New value: +"Refuse before STT when estimatedCredits would exceed this (400 cost_exceeds_max, 0 credits)."
    • changedInput schema / properties / url / description
      Previous value: -"Public YouTube video URL, e.g. https://youtube.com/watch?v=ID. Not a TikTok/Instagram/Facebook URL. The URL platform must match this tool's platform. Do not pass cross-platform URLs, e.g. YouTube to TikTok, Instagram to Facebook, LinkedIn to X/Twitter, or Pinterest to Rumble."New value: +"Public YouTube video URL, e.g. https://youtube.com/watch?v=ID. Not a TikTok/Instagram/Facebook URL. The URL platform must match this endpoint's platform. Do not pass cross-platform URLs, e.g. YouTube to TikTok, Instagram to Facebook, LinkedIn to X/Twitter, or Pinterest to Rumble."
  7. Added

TDQS

A4.4/5.0
Behavior4/5

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

With no annotations, the description carries the full burden and mostly succeeds: it discloses pricing (2 credits per started audio minute), no charge on empty results or failures, default-fresh vs cached behavior, and the 400 cost_exceeds_max refusal path. It does not describe the transcript payload (segments, timestamps, language handling) or any rate limits, which leaves a modest gap for an unannotated tool with no output schema.

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?

Front-loaded with the purpose, then usage condition, then cost, then the cache tip — each sentence adds new information. No restatement of the name or filler.

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?

Covers the risk profile (cost, failure billing, refusal path) well for a paid STT tool, which is what an agent most needs before calling. The main omission is return-shape guidance — with no output schema, the agent gets no sense of whether the transcript includes segments, timestamps, or detected language.

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

Parameters3/5

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

Schema description coverage is 100%, so the schema already documents url, cache, fields, compact, language and maxCredits. The description reinforces the cost consequence of cache=true and of exceeding maxCredits, but adds no syntax or format detail beyond the schema, and says nothing about fields/compact/language. Baseline 3 applies.

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?

States a specific verb+resource ("Speech-to-text for YouTube audio") and immediately distinguishes the output from YouTube's published captions, which is exactly the axis that separates it from youtube_transcript. An agent can tell what it gets without opening the schema.

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

Usage Guidelines5/5

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

Gives explicit positive conditions ("when a video has no captions — or when you want a transcript of what was actually spoken rather than YouTube's published captions"), which implicitly routes caption-reading cases to the sibling transcript tool. No alternative is named outright, but the decision rule is unambiguous.

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

Try in Browser

Glama MCP Gateway

Add one secure layer between your agents and this server.