Video To Text

Video To Text is a video to text converter api that converts a public video URL from a supported source or an authorized media upload into detected-language text, ordered timestamps, source metadata, and optional translation for data and automation workflows.

• One successful input writes one Dataset item with 22 top-level fields.
• The transcript contains full text and ordered start, end, and text segments.
• The optional translation selector exposes 133 target languages.
• The public candidate completed in 19.503 seconds on August 12, 2026 and produced one Dataset item.

Run a one-result test · View API

Start with https://kick.com/kick_clipz/clips/clip_01KZH2QVDSPYWHRPFK52475GZJ and no translation to test the smallest path. The FREE-tier fixed events are $0.53 for one transcript plus a 1 GB start, followed by metered Actor usage.

What does Video To Text do?

Video To Text downloads the accessible media behind a public video URL from a supported source or an authorized media upload, recognizes its speech, and publishes one normalized record with source context and timestamped text. Use one normalized transcript shape across supported public video sources or an authorized uploaded media file.

The result keeps recognition in transcript and, when requested, a second aligned object in translation. Each object contains a language label, combined text, and ordered segments. Source metadata can include title, description, author, duration, thumbnail, publication time, audio attribution, taxonomy, and engagement counts when the page exposes them.

This Actor does not search supported public video, open private content, label speakers, read text drawn inside video frames, create a video download product, or decide whether you have legal rights to use the media. YouTube's official sharing help describes the source's own sharing or media behavior.

How do I run Video To Text?

Run Video To Text with one public media item and inspect the single Dataset row before scheduling a larger workflow.

1. Open the Actor input and paste https://kick.com/kick_clipz/clips/clip_01KZH2QVDSPYWHRPFK52475GZJ into video_url.
2. Leave translate empty for the original detected language, or choose spanish to reproduce this page's translated scenario.
3. Provide video_url or video_file; the uploaded file wins when both are present.
4. Start the Actor and wait for a terminal status.
5. Open the default Dataset and confirm that transcript.text and transcript.segments match the source audio.

{
  "video_url": "https://kick.com/kick_clipz/clips/clip_01KZH2QVDSPYWHRPFK52475GZJ",
  "translate": "spanish"
}

The billing unit is one completed transcript row. A failed source retrieval or silent-media path writes no result row and therefore does not represent a successful transcript event.

What data does Video To Text return?

Video To Text returns one 22-field Dataset item combining processing details, source metadata, engagement values, recognized speech, and optional translation.

Data group	Fields
Processing	processor, processed_at, platform
Source identity	title, description, author, author_id, author_url, duration, published_at, thumbnail
Audio and taxonomy	audio_title, audio_artist, categories, tags
Engagement	view_count, like_count, shares_count, dislike_count, comment_count
Speech	transcript, translation

The item below is abbreviated to show the nested speech shape. A successful real item retains all 22 top-level fields and all returned segments.

{
  "platform": "supported public video",
  "title": "Source title when available",
  "author": "Source creator when available",
  "duration": 24.4,
  "transcript": {
    "language": "English",
    "text": "Abbreviated recognized speech.",
    "segments": [
      {
        "start": "00:00:00.000",
        "end": "00:00:02.400",
        "text": "Abbreviated recognized speech."
      }
    ]
  },
  "translation": {
    "language": "Spanish",
    "text": "Traducción abreviada.",
    "segments": [
      {
        "start": "00:00:00.000",
        "end": "00:00:02.400",
        "text": "Traducción abreviada."
      }
    ]
  }
}

Unavailable source values remain null rather than being converted to zero or guessed. The Dataset is exportable in Apify-supported formats; JSON preserves the nested transcript and translation structures most directly.

What inputs can I configure?

Video To Text exposes 3 inputs in schema order, with one URL requirement and one optional translation selector plus an upload alternative.

Parameter	Type	Required	Description	Example
video_url	string	No	One public media page URL in this Actor's documented scope.	https://kick.com/kick_clipz/clips/clip_01KZH2QVDSPYWHRPFK52475GZJ
video_file	string	No	An authorized audio or video file up to 8 GB; when supplied, it takes priority over video_url.	Upload or direct media-file URL
translate	string	No	Optional target language; the schema exposes 133 choices.	spanish

translate is optional; omitting it does not imply English or any other target. The recognition path detects the source language, while a selected target asks for a separate translated object. Provide video_url or video_file; the uploaded file wins when both are present.

What platforms and markets does Video To Text cover?

Video To Text covers a public video URL from a supported source or an authorized media upload rather than a country, market, account, feed, search page, or URL batch.

The platform-specific boundary makes integration tests clear: use a direct media item that a normal public visitor can open. Profiles, channel pages, playlists, searches, private shares, removed posts, credential-gated media, unrelated domains, and live pages without a retrievable recording are outside the input boundary.

Language coverage is separate from platform coverage. Speech recognition detects the source language, and translate offers 133 target choices. Source page availability, regional rules, and the clarity of the audio can still affect results.

Why use Video To Text?

Use Video To Text when the workflow needs structured, time-addressable speech rather than manual playback or an unstructured text file.

Capability	Workflow benefit
One public media URL per run	Small, reproducible jobs with a clear source identity
Ordered transcript segments	Search a phrase and return to the corresponding media time
22 stable top-level fields	One downstream reader for speech and source context
133 translation targets	Keep original recognition and translated text together
Apify Dataset and API	Export, schedule, webhook, or connect the result without managing storage

The main trade-off is that transcription requires media retrieval and speech processing. It is heavier than reading source-supplied captions alone, and accuracy can decline with noise, music, accents, overlapping speakers, or specialized names.

Who is Video To Text for?

Video To Text is for developers, data teams, researchers, accessibility workflows, and content operations that already know the public media URL they are authorized to process.

Developers can call one Actor and receive predictable JSON. Researchers can make spoken material searchable while retaining source metadata. Content teams can find quotes or draft notes from time-coded segments. Multilingual reviewers can compare transcript and translation without losing segment order. Use one normalized transcript shape across supported public video sources or an authorized uploaded media file.

It is not the right tool for private-content access, account automation, speaker diarization, video discovery, continuous live captioning, bulk URL search, subtitle-file editing, or rights clearance.

How can I use Video To Text through the API or MCP?

Use public Actor ID E9f5oS7cOn2Kgw0uy or the readable name truefetch/video-to-text with the same one-item JSON used in Console.

$curl -X POST "https://api.apify.com/v2/acts/E9f5oS7cOn2Kgw0uy/run-sync-get-dataset-items?token=$APIFY_TOKEN"   -H "Content-Type: application/json"   -d '{"video_url": "https://kick.com/kick_clipz/clips/clip_01KZH2QVDSPYWHRPFK52475GZJ", "translate": "spanish"}'

The synchronous endpoint is convenient for a short item. For longer media, start an asynchronous run, poll its status, and read the default Dataset after success. Keep the Apify token in an environment variable or secret store rather than source code.

For MCP, configure the official Apify MCP endpoint and expose truefetch/video-to-text as an Actor tool. The live API page provides generated examples for supported clients and the OpenAPI definition.

How much does Video To Text cost?

On the FREE tier, one transcript is $0.52, the default 1 GB Actor Start is $0.01, and the smallest successful run has $0.53 in fixed events plus metered usage.

Tier	Transcript	Translation	Actor Start	Actor usage
FREE	$0.52	$0.18	$0.01 per GB, minimum one event	$0.00001 per usage event
BRONZE	$0.50267	$0.174	$0.01 per GB, minimum one event	$0.00001 per usage event
SILVER	$0.48533	$0.168	$0.01 per GB, minimum one event	$0.00001 per usage event
GOLD	$0.468	$0.162	$0.01 per GB, minimum one event	$0.00001 per usage event
PLATINUM	$0.468	$0.162	$0.01 per GB, minimum one event	$0.00001 per usage event
DIAMOND	$0.468	$0.162	$0.01 per GB, minimum one event	$0.00001 per usage event

The Transcript event is charged for one completed recognized result. Translation adds $0.18 on the FREE tier when requested and completed. For the example with Spanish, the fixed-event total is $0.71 plus usage. Actor usage varies with media length, retrieval, and processing time. Confirm the live pricing page before relying on a rate because prices can change.

How does Video To Text compare with alternatives?

Video To Text is the best fit when one known public supported public video item must become normalized JSON with timestamped speech and optional translation.

Option	Best fit	Trade-off
Video To Text	One public source URL, source metadata, and time-coded text	Requires a retrievable public media item and bills a run
Manual playback and typing	A very short, one-time review	Difficult to timestamp, repeat, schedule, or export consistently
The source's own caption or transcript feature	The source already exposes authorized text in the needed format	Availability and export shape depend on the source and media owner
A general file transcription category	You already possess an authorized local file	Source URL context and public-page metadata may be absent

Choose the universal Video To Text edition when uploads or mixed supported sources matter. Choose this edition when supported public video scope is the acceptance criterion and unrelated URLs should fail early.

What are the limits and troubleshooting steps?

The main limits are public access, source extraction, audio quality, media duration, and the one-item-per-run input shape.

• No row is produced: confirm https://kick.com/kick_clipz/clips/clip_01KZH2QVDSPYWHRPFK52475GZJ still opens publicly and contains playable media, then retry without translate.
• A private or removed page fails: make the media publicly accessible through the source's normal controls or use an authorized local file with the universal Actor; do not provide credentials.
• Some metadata is null: the source did not expose that value. Preserve null rather than treating it as zero.
• Names or jargon are wrong: compare the affected segment timestamp with the audio and correct the downstream text before publication.
• Translation is absent: first confirm that the original transcript succeeded, then retry one target. A partial translation is not attached.
• A long item runs slowly: start with the short candidate to isolate URL handling, then test representative duration and cost before automation.

For a reproducible report, open an Actor Issue and include the run ID, exact public input, translation target, expected behavior, and the first relevant log line. Do not attach private media or secrets.

Frequently asked questions

universal video to text api?

Yes for a supported public item: the Actor needs an Apify token, not a developer key for supported public video, and returns timestamped Dataset JSON.

bulk universal transcript export?

Run one Actor execution per URL, then export or combine Dataset items through your orchestrator. Each successful URL is its own transcript billing unit.

Does Video To Text return timestamps?

Yes. transcript.segments contains ordered start, end, and text values, and a completed translation follows the same order.

Can Video To Text process private content?

No. The Actor is for public or otherwise directly authorized media and does not accept account credentials to bypass source visibility.

Can I upload a file instead of a URL?

Yes. video_file accepts an authorized media upload or direct file URL up to 8 GB and takes priority over video_url.

Why are some values null?

Null means a source value was unavailable. It does not mean zero views, zero likes, or an empty creator identity.

Can I schedule the same media URL?

Yes. Apify schedules can repeat the input, but every execution is separately billed and source availability or metadata may change.

Does a translation replace the original transcript?

No. The original remains in transcript, while the requested language is stored separately in translation after all segments complete.

Related TrueFetch Actors

The closest TrueFetch choices separate universal media input from platform-specific transcript boundaries.

• Instagram Video Transcript — a platform-specific transcript workflow
• Snapchat Video Transcript — a platform-specific transcript workflow
• Linkedin Video Transcript — a platform-specific transcript workflow

Support

Ask about Video To Text, its input boundary, or Dataset fields in the TrueFetch community on Telegram. For a reproducible defect, use the Actor Issues page and include the run ID plus the exact non-sensitive input.

Last Updated: August 12, 2026