transcriptfetchGet API access

Video Transcript API

One Video Transcript API for YouTube, TikTok, Instagram and podcasts

Get the transcriptTry in ChatGPT or Claude
100 free credits · no card

YouTube, TikTok, Instagram, podcasts and direct media files, through one endpoint that returns the same JSON for all of them. Captions when the source has them, transcription when it does not. One credit per successful response, nothing at all when a fetch fails.

mixed sources · 4 fetched200 OK
00:00youtube.com/watch?v=dQw4w9WgXcQ — captions, 412 segments
00:00tiktok.com/@dev/video/7137723462 — transcribed, 8 segments
00:00instagram.com/reel/DbSfnc9TZ7q — transcribed, 6 segments
00:00open.spotify.com/episode/4rOoJ6Eg — transcribed, 2,184 segments
One request shape. One response shape. One credit each.
captions + transcribed · en1 credit

Every platform breaks differently.

One integration covers all of them. Adding a source is a different URL, not a new library, a new auth flow and a new set of failures.

One library per platform

A caption package for YouTube, a scraper for TikTok, the Graph API for Instagram, and something else again for podcast audio. Four failure modes, four sets of proxies, and four response shapes to normalise before your own code can start.

With TranscriptFetch

One endpoint, one bearer token, one envelope. Captions where they exist, transcription where they do not, and the same segments array either way. Platform changes are ours to chase, and failed requests are never charged.

Who it is for, and how they use it

Three ways in, depending on how you work. Same endpoint, same credit, same response shape underneath all of them.

DevelopersOne endpoint

Engineers building a product

Your users paste links from anywhere, and you do not want a branch in your code for each platform.

  1. 01Create a key in the dashboard, send it as a Bearer token.
  2. 02POST whatever URL the user gave you, unparsed.
  3. 03Read the same segments array whatever the source was.
Read the API reference →
RAG and pipelinesRun at volume

Teams building on retrieval

Half your source material is spoken, and none of it is searchable until it becomes text with timestamps.

  1. 01Send up to 50 mixed URLs in one batch call.
  2. 02Chunk on the segment boundaries you get back.
  3. 03Cite the source and the second, not just the document.
See batch fetching →
AI and agentsNo code

Teams building with assistants

Your agent gets handed a link and should not need to know which platform it came from.

  1. 01Add the MCP server once, no deploy and no glue code.
  2. 02Ask in plain language, the assistant calls the tool itself.
  3. 03Answers come back with the timecode attached.
Set up MCP →

One request, whichever way you build.

Write code if you want to. If you would rather not, the same fetch runs from a workflow tool or straight inside an assistant.

REST

Call the endpoint

One POST with any supported URL. Same JSON for all of them.

Read the API reference →
NO CODE

Drop it in a workflow

An official n8n node. Run it on a schedule or a trigger.

See the n8n node →
MCP

Let an assistant do it

Connect once and Claude or ChatGPT fetches transcripts mid-chat.

Set up MCP →
curl -X POST https://transcriptfetch.com/api/v1/transcripts/video \
  -H "Authorization: Bearer $TRANSCRIPTFETCH_KEY" \
  -d '{"video": "https://www.youtube.com/watch?v=dQw4w9WgXcQ"}'

# the same call, a different URL:
#   tiktok.com/@dev/video/713772346
#   instagram.com/reel/DbSfnc9TZ7q
#   open.spotify.com/episode/4rOoJ6Eg
No code required

Let the assistant fetch the video itself.

Add the MCP server once and Claude, ChatGPT or Cursor can pull a transcript mid-conversation, from any supported platform. Same key, same credit, no glue code and nothing to deploy.

Set up MCPOr browse the n8n node →

https://transcriptfetch.com/mcp

ClaudeTool: transcriptfetch
Summarise this video and quote the part about retrieval.
CALLING get_transcript(video="youtube.com/watch?v=dQw4…")
At 00:23 she says retrieval moved the numbers, not model size.

Questions people ask first

One transcript API vs one library per platform

TranscriptFetch
One library per platform
Coverage
YouTube, TikTok, Instagram, podcasts and direct media files
A caption package, a scraper, the Graph API and an audio pipeline, one each
Adding a source
A different URL, same call
A new library, a new auth flow, a new set of failures
Response shape
One envelope, the same segments array everywhere
Four shapes to normalise before your code starts
No captions on the source
Transcribed with AI speech-to-text, same endpoint
Depends on the platform; often nothing
Blocks and proxies
Handled for you, rotating infrastructure
Yours to run, per platform
Batch
Up to 50 mixed URLs in one call
Manual loops per platform
Failed fetches
Free, billed on delivery only
You pay for the compute either way
MCP server
Yes, for Claude, Cursor and others
No

What happens when a source has no captions

TranscriptFetch
A caption-only fetch
Source has no caption track
Transcribed with AI speech-to-text
Request fails or returns nothing
Changes needed in your code
None, same video endpoint
Separate fallback logic per platform
Extra cost
Charged only when a transcript is delivered
Premium tier, or unavailable
Blocks and rate limits
Handled by our infrastructure
Yours to solve

What the endpoint returns

One response shape across every supported platform
Captions when the source has them, AI transcription when it does not
Per-segment timestamps: start and duration
Batch endpoint for up to 50 mixed URLs per call
Plain text and segmented JSON in one response
Channels, playlists and keyword search on YouTube
Webhooks or polling for transcription jobs
An MCP server, plus Python and JavaScript SDKs

Guides from the blog

How to transcribe a video: every method, honestly comparedFive ways to turn any video into text: platform captions, free online tools, AI transcription APIs, offline Whisper, and when manual is stil…Read the guide →How to repurpose TikTok videos, starting from the transcriptMost repurposing advice tells you to repost the video somewhere else. The higher-leverage move is to pull the words out first: one TikTok tr…Read the guide →

Turn videos into structured text at scale

Our Video Transcript API gives you a reliable way to extract spoken content from video files and hosted media, then turn it into clean, usable transcript data for your product, workflow, or internal tools.

Whether you are building search, summaries, captions, compliance systems, or content intelligence pipelines, the API is designed to help you move from raw video to structured text with minimal integration work.

What you can do with the API

  • Extract transcripts from video content for downstream analysis, archiving, or accessibility workflows
  • Process video at scale with an API-first workflow built for automation
  • Convert spoken dialogue into machine-readable text that can be indexed, searched, or transformed
  • Use transcript output in your own applications for analytics, moderation, support, media operations, or knowledge management
  • Integrate transcription directly into existing systems without building speech processing infrastructure yourself

Core features

Video-to-text transcription

Submit video content and receive transcript output that represents the spoken audio as text. This makes it easier to work with interviews, webinars, training videos, product demos, meetings, lectures, and other spoken-content formats.

Structured output for developers

Transcript results are designed to be easy to parse and use in applications. You can feed output into search systems, summarization pipelines, subtitle generators, internal dashboards, or document stores.

API-first integration

The service is built for developers who want a straightforward way to add transcription capabilities to products and automations. Use the API as part of batch pipelines, user-facing features, or backend media processing workflows.

Scalable processing

Handle anything from individual files to larger transcription workloads. The API supports teams that need a repeatable, programmatic way to process video content without manual intervention.

Useful for many product workflows

Transcript data can support a wide range of use cases, including:

  • Search and discovery across video libraries
  • Summaries and highlights for long-form content
  • Captions and accessibility experiences
  • Compliance and recordkeeping for regulated environments
  • Knowledge extraction from internal or customer-facing media
  • Content moderation and review workflows
  • Training and documentation generation from spoken material

Built for production use

For teams shipping real products, transcript infrastructure needs to be dependable and easy to operationalize. The Video Transcript API helps reduce the complexity of handling speech extraction so your team can focus on product logic, user experience, and downstream intelligence.

Common use cases

  • Media platforms that need searchable video libraries
  • Edtech products converting lessons into readable study material
  • Enterprise teams indexing internal recordings and training content
  • Support organizations extracting insights from recorded calls or walkthroughs
  • Research and analysis teams turning video archives into queryable text
  • Accessibility workflows that need transcript generation for end users

Why developers use it

  • Fast to integrate
  • Useful across many video formats and workflows
  • Easy to connect to existing pipelines
  • Designed for applications that need transcript data as a building block

If you need to transform video content into text that your systems can understand and use, the Video Transcript API provides a simple foundation for doing it programmatically.

Start with 100 free credits

Paste any supported link or copy the curl, whichever suits you. No card, no sales call.

Read the docsGet started free