Extract full closed captions, timed transcript lines, and complete text from YouTube videos through MuAPI. Extract complete closed captions, subtitles, and transcript text from any YouTube video (3x SERP).
Extract complete closed captions, subtitles, and transcript text from any YouTube video (3x SERP).
About this model
YouTube Video Subtitles extracts the full closed-caption track from a video, returning both a timed line-by-line breakdown (with start and end offsets in seconds) and a single concatenated full_text transcript. This removes the need to manually download and parse caption files, and works for both creator-uploaded captions and YouTube's auto-generated ones where available. The timed structure makes it straightforward to jump to a specific moment referenced in the transcript, while the flat text is ready to feed directly into a summarization or topic-extraction pipeline.
Cost analysis
| Provider | Cost | Notes |
|---|---|---|
| muapiapp | From $0.010 per Live lookup | Subtitles are priced at 3x the selected SERP method; Standard normal/high starts at $0.003/$0.006. |
| Fal.ai | Not available | Not offered as a standalone API. |
| Replicate | Not available | Not offered as a standalone API. |
Subtitles are priced at 3x the selected SERP method; Standard normal/high starts at $0.003/$0.006.
Not offered as a standalone API.
Not offered as a standalone API.
** Competitor pricing is estimated based on similar model architectures and usage tiers.
Configuration schema
| Parameter | Type | Description | Default |
|---|---|---|---|
| Video Id | string | YouTube video ID or full YouTube URL (watch, embed, short link). | dQw4w9WgXcQ |
| Subtitles Language | string | Language tag of subtitles to extract (e.g., 'en', 'es', 'fr', 'zh-Hans'). | en |
| Subtitle Translation Language | string | Optional language tag for translating the selected subtitle track. | null |
| Location | Enum (213 options) | Country to target for the search. | United States |
| Language | Enum (129 options) | Language to search in. | English |
| Operating System | Enum (2 options) | Desktop operating system. | null |
| Tag | string | Optional request tag for tracing. | null |
| Delivery Method | Enum (2 options) | Use Live for immediate results or Standard for queued tasks. | live |
| Priority | Enum (2 options) | Standard task priority. | normal |
YouTube video ID or full YouTube URL (watch, embed, short link).
dQw4w9WgXcQLanguage tag of subtitles to extract (e.g., 'en', 'es', 'fr', 'zh-Hans').
enOptional language tag for translating the selected subtitle track.
nullCountry to target for the search.
United StatesLanguage to search in.
EnglishDesktop operating system.
nullOptional request tag for tracing.
nullUse Live for immediate results or Standard for queued tasks.
liveStandard task priority.
normalDeveloper documentation
video_id.subtitles_language, such as en, es, or zh-Hans (defaults to en).subtitles_translate_language when the requested caption track should be translated.start, end, text) and the full concatenated transcript text.curl -X POST https://api.muapi.ai/api/v1/seo-youtube-video-subtitles \
-H "x-api-key: YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"video_id": "dQw4w9WgXcQ",
"subtitles_language": "en"
}'
Frequently asked
The API returns an empty list of subtitles, or falls back to available auto-generated captions where they exist.
Yes. Each item in the `items` array includes `start` (seconds), `end` (seconds), and `text`.
Yes, the response includes both the `items` array of timed lines and a `full_text` field with the entire transcript concatenated.
Yes, it retrieves whichever caption track is available for the requested language, whether creator-uploaded or auto-generated by YouTube.
Yes, the `full_text` field is designed to be passed directly into a summarization or topic-extraction LLM prompt.