YouTube Transcript Scraper Pro (Captions + AI Fallback)
Pricing
from $0.70 / 1,000 transcript extracteds
Extract YouTube transcripts at scale without burning through your budget. It starts with free captions whenever they're available, then switches to AI only for videos that don't have them. You stay in control of costs, and the output — JSON, SRT, VTT, plain text, or LLM-ready format
url
Optional
Input URL of the video (present on extraction-error items where metadata could not be fetched).
metadata
Optional
Video metadata fetched via yt-dlp. Null on extraction-error items.
{ "id": "dQw4w9WgXcQ", "title": "Example Video Title", "channel": "Example Channel", "duration": 213}Language
language
Optional
Language code of the transcript. For native YouTube captions this is any of the 130+ YouTube caption codes (e.g. en, zh-TW, pt). For AI transcriptions this is one of the 99 codes supported by the Whisper model (e.g. en, zh, de). See the Language Reference section in the README for full lists.
Auto-generated
is_auto_generated
Optional
True if YouTube auto-generated the captions.
AI Transcribed
is_ai_generated
Optional
True if the transcript was produced by the built-in speech-to-text model.
transcript_json
Optional
Timestamped transcript segments. Each segment has start (seconds), end (seconds), and text. When wordLevel is enabled, each segment also includes a words array with per-word timestamps: [{start, end, text}].
[ { "start": 0, "end": 3.5, "text": "Hello world.", "words": [ { "start": 0, "end": 1.2, "text": "Hello" }, { "start": 1.3, "end": 3.5, "text": "world." } ] }]Plain Text
transcript_text
Optional
Full transcript as plain text, segments joined with spaces.
LLM-Ready Text
transcript_llm
Optional
LLM-optimised transcript: filler tokens ([Music], (laughter), etc.) stripped, whitespace normalised.
SRT Subtitles
transcript_srt
Optional
Transcript in SRT subtitle format.
WebVTT Subtitles
transcript_vtt
Optional
Transcript in WebVTT format.
Language Confidence
language_probability
Optional
AI model's confidence in the detected language (AI transcription only, 0–1).
Language Forced
language_was_forced
Optional
True if the language was set via forceWhisperLanguage (AI transcription only).
AI Min Charged
ai_duration_charged_min
Optional
Number of AI minutes charged for this video (AI transcription only).
Speech Duration (sec)
ai_speech_duration_sec
Optional
Actual speech duration detected by the AI model in seconds (AI transcription only).
Available Languages
available_languages
Optional
Only present on NO_CAPTIONS_AVAILABLE and LANGUAGE_NOT_FOUND error items. Lists the caption language codes YouTube does provide on this video — use them to refine your languages input.
[ "en", "es", "fr"]Error Message
error
Optional
Human-readable error message when transcript extraction failed.
error_code
Optional
Structured error code for programmatic filtering. Error items are never billed — charges only apply to successful transcripts. Input errors: AGE_RESTRICTED (requires sign-in / age verification), PRIVATE_OR_UNAVAILABLE (private, deleted, or region-blocked), LIVE_VIDEO (live streams have no static captions), LANGUAGE_NOT_FOUND (captions exist but not in the requested language). Budget errors: NO_CAPTIONS_AVAILABLE (video has zero caption tracks), AI_MINUTES_LIMIT_REACHED (maxAiMinutes budget exhausted), AI_FALLBACK_SKIPPED_TOO_LONG (video exceeds skipAiFallbackIfLongerThan), SPENDING_LIMIT_REACHED (Apify account spending limit hit). Infrastructure errors: BOT_DETECTION (YouTube challenged the request, retried via proxy), EXTRACTION_ERROR (generic yt-dlp failure), AI_TRANSCRIPTION_FAILED (Whisper model or audio download error), UNEXPECTED_ERROR (unhandled exception).