TED Talk Downloader

TED Video Downloader turns one public TED Talk page into structured media metadata and, when requested, a video file stored with the run.

• One URL keeps each record traceable to a specific talk page.
• Metadata mode supports lightweight cataloging before any file transfer.
• The Dataset presents artwork first and preserves a 26-key cross-video shape.
• Comments are outside the TED contract, so no comment limit is advertised.

Prepare a test run

Read the Apify API reference

Publish the Actor and replace .

Why Choose This API

A TED Talk page combines editorial text, speaker attribution, artwork, timing, and playable media. This Actor translates the available pieces into a normalized row suitable for a research index, learning catalog, review queue, or downstream media workflow.

The public contract starts from a known talk. It does not search TED, enumerate playlists, crawl a series, collect transcripts, or discover related speakers. Although the bundled extractor library recognizes additional TED page types, this Actor deliberately promises only an individual public Talk page so its Store description matches a testable unit.

metadata is the prudent first pass: it verifies identity and availability without storing a video. The other three values ask for an available rendition. They cannot create 4K media when the source provides less.

Quick Start Guide

Use the following URL-shaped sample to understand the request:

{
  "video_url": "https://www.ted.com/talks/ken_robinson_do_schools_kill_creativity",
  "video_quality": "metadata"
}

example_talk is intentionally not presented as a real talk. Replace it with the canonical page for a public TED Talk you may process. Confirm the page loads and plays in a private browser session, then compare the Actor’s title, speaker, thumbnail, ID, and duration with the source.

Do not substitute a TED search result, topic page, series, playlist, event landing page, or embedded player in release testing. If a media copy is needed, request it only after the metadata result points to the intended talk.

Input Parameters

Name	Kind	Required	Accepted meaning
video_url	string	Yes	Canonical, publicly accessible TED Talk page.
video_quality	string	No	metadata, low, medium, or high; omission selects metadata behavior.

The required field carries the same placeholder in prefill and example. The optional selector has no preselected value or example, preserving the caller’s explicit intent.

No transcript language, speaker, topic, event, playlist, login, cookie, comment, search, or batch option is exposed. The schema rejects undeclared properties.

Output Data Schema

One result may contain 39 fields arranged for review:

Concern	Keys
Talk identity	thumbnail, id, platform, title, description, duration, published_at
Speaker/source	author, author_id, author_url
Classification and metrics	categories, tags, view_count, like_count, shares_count, dislike_count, comment_count
Technical media	width, height, fps, audio_title, audio_artist, video
Processing context	comments, processed_at, processor

Illustrative output for :

{
  "id": "example_talk",
  "platform": "TedTalk",
  "title": "Illustrative TED Talk",
  "author": "Illustrative speaker",
  "thumbnail": "https://example.com/ted-talk-cover.jpg",
  "comments": [],
  "video": ""
}

The values above are documentation, not observations from TED. Some general video counters or audio fields may be null because the Talk page does not supply them. An empty video is expected for metadata-only work. The reserved comments array does not imply a TED comment collector.

Source-dependent fields also present in every item:

Field	Meaning
source_url	Canonical page URL of the item on its own platform
media_type	Kind of media the source reports, such as video or audio
live_status	Whether the item is a recording, currently live, or a finished stream
availability	Source visibility, such as public, unlisted, or needs_auth
age_limit	Minimum viewer age the source enforces; zero means no restriction
followers	Follower or subscriber count of the author at processing time
is_verified	Whether the author's account carries a verified badge
saves_count	Times viewers saved the item, on sources that expose it
file_ext	Container extension of the selected media, such as mp4 or webm
file_size	Size of the selected media in bytes, exact or approximate
chapters	Chapter markers with title and start and end time in seconds

Integration Examples

The intended handle is agentx/ted-video-downloader. It currently has no Store-backed ID. A future HTTP call for the same fictional scenario would look like this:

curl -X POST "https://api.apify.com/v2/acts/agentx~ted-video-downloader/runs?token=$APIFY_TOKEN" \
 -H "Content-Type: application/json" \
 -d '{"video_url":"https://www.ted.com/talks/ken_robinson_do_schools_kill_creativity","video_quality":"metadata"}'

The equivalent MCP configuration retains those arguments:

{
  "mcpServers": {
    "ted-talk-media": {
      "url": "https://mcp.apify.com?tools=agentx/ted-video-downloader",
      "headers": {
        "Authorization": "Bearer <APIFY_TOKEN>"
      }
    }
  },
  "exampleToolArguments": {
    "video_url": "https://www.ted.com/talks/ken_robinson_do_schools_kill_creativity",
    "video_quality": "metadata"
  }
}

Keep credentials outside source control. The Apify MCP documentation explains hosted authentication and tool exposure.

Pricing & Cost Calculator

The repository declares the following local event schedule; it is not yet a public Store offer:

Event key	Local definition	Amount
actor_usage	Runtime, proxy, and storage metering	$0.00001
metadata	One processed talk without saved video	$0.05
download	One processed talk with saved video	$0.22

On those declarations, the minimum metadata example is $0.05 plus metered usage, while a download begins at $0.22 plus metered usage.

Use Cases & Applications

Learning teams can register authorized talks in a course catalog. Researchers can attach normalized source metadata to citations. Editors can queue a permitted video for review. Accessibility or language systems can hand an allowed file to a separate transcription or caption workflow.

For change tracking, retain each dated Dataset item and compare outside the Actor. This tool neither watches TED nor discovers new talks.

Limits & Troubleshooting

TED can remove a talk, change its page, restrict playback, provide only certain renditions, or alter embedded media delivery. Geographic access, network throttling, and extractor changes may also affect a run. There are no login, cookie, or region inputs.

When a request fails, first replace the placeholder and open the canonical Talk page privately. Confirm the page is an individual talk rather than a collection. Retry using metadata. If metadata is correct but storage is empty, check the requested quality, logs, and key-value store.

Release regression should cover a current public Talk, an event or playlist URL expected to be rejected by the documented scope, a removed page, and one authorized file run. Record the extractor key, talk ID, thumbnail, mode, final status, and stored-file result.

FAQ

Will this search TED by topic or speaker?

No. Supply the exact talk page yourself.

Does it return the official transcript?

Transcript retrieval is not part of this Actor’s inputs or output promise.

Can it download a whole TED series?

No. The public unit is one Talk page per run.

Are comments collected?

No comment control exists, and the compatibility array is normally empty.

Is the selected resolution guaranteed?

No. The quality value is constrained by renditions TED actually exposes.

Why might metrics be null?

The normalized model includes cross-platform keys that an individual TED page may not publish.

Can a scheduler rerun a talk?

Yes; each scheduled run processes the supplied page again.

Does the returned language field describe the audio or the subtitles?

language reports the primary spoken language TED publishes for the talk. Translated caption tracks are listed separately in subtitles as language codes. A talk can therefore report English audio while offering dozens of subtitle codes, and neither field implies that a translated audio track exists.

Legal & Responsible Use

Download or retain a Talk only when you have the necessary rights. Callers remain responsible for TED terms, speaker and publisher rights, copyright, privacy, applicable law, and retention decisions.

Related Tools

• Video Transcript creates text from permitted media.
• Video Captions Downloader retrieves available caption tracks.
• Video to Social Post helps draft from authorized source media.
• All Video Scraper provides a general multi-site interface.

Support & Community

The 404 Store result means there is no verified Actor ID, public review record, issue history, usage count, or Store run to cite yet.

• AgentX community
• Message AgentX
• Create or open an Actor
• REST API docs
• MCP integration docs

After launch, provide the stable ID, redacted Talk URL, exact input, and run ID when requesting support.

Prepare a test run

Read the Apify API reference