Instagram Transcript Scraper

Turn Instagram videos and reels into clean, structured transcripts with ease.

Extract spoken content from public Instagram videos — designed for developers, data teams, and AI agents that need text, not just media.

Overview

The Instagram Transcript Scraper extracts transcripts from Instagram video posts and reels using a dual-strategy approach:

1. Native Captions — Extracts Instagram's auto-generated captions when available (fastest, zero extra cost)
2. Whisper AI — Falls back to OpenAI's Whisper speech-to-text model for accurate transcription of any video with speech

What You Get

For each Instagram video URL, the scraper returns:

• Full transcript text (complete speech-to-text)
• Timestamped segments for precise alignment
• Video title & thumbnail
• Video and audio playback URLs
• Engagement metrics (likes, comments)
• Creator profile information (username, display name, avatar)
• Transcription method used (native or whisper)
• Error status per request (safe for automation)

All data is returned in structured JSON, ready to feed into databases, AI pipelines, or downstream services.

Input Parameters

Field	Type	Required	Description
videoUrls	array	Yes	List of Instagram video/reel URLs to transcribe
transcriptionMethod	string	No	auto (default), native, or whisper
whisperModel	string	No	tiny, base (default), or small
language	string	No	Language code (e.g. en, es). Empty = auto-detect

Input Example

{
  "videoUrls": [
    "https://www.instagram.com/reel/DGtml1wNRoM/",
    "https://www.instagram.com/p/DULBkEngpxg/"
  ],
  "transcriptionMethod": "auto",
  "whisperModel": "base"
}

Output Format

Each transcript segment is returned as a separate dataset row, with full video metadata included on every row for easy filtering and export.

Field	Type	Description
url	string	Normalized Instagram video URL
code	string	Instagram shortcode
pk	string	Instagram internal post ID
id	string	Combined media identifier
title	string	Video caption text
img	string	Video thumbnail URL
videoUrl	string	Direct video playback URL
audioUrl	string	Direct audio playback URL
createTime	number	Video creation timestamp (Unix seconds)
likeCount	number	Number of likes
commentCount	number	Number of comments
userPk	string	Creator user ID
userName	string	Creator username
userFullName	string	Creator display name
avatarUri	string	Creator avatar image URL
fullText	string	Complete transcript text
segmentIndex	number	Segment index (0-based)
segmentStart	number	Segment start time in seconds
segmentEnd	number	Segment end time in seconds
segmentText	string	Text for this specific segment
totalSegments	number	Total number of segments for this video
transcriptionMethod	string	Method used: native or whisper
errMsg	string	Error message (empty if success)
timestamp	string	Response timestamp (ISO 8601)

Output Example

{
  "url": "https://www.instagram.com/p/DTxiM0Ijqvz/",
  "code": "DTxiM0Ijqvz",
  "pk": "3814980773552761843",
  "id": "3814980773552761843_48143082417",
  "title": "#explore",
  "img": "https://scontent-...",
  "videoUrl": "https://scontent-...",
  "audioUrl": "https://scontent-...",
  "createTime": 1769001195,
  "likeCount": 171650,
  "commentCount": 781,
  "userPk": "48143082417",
  "userName": "theavamariee",
  "userFullName": "Ava Marie",
  "avatarUri": "https://scontent-...",
  "fullText": "Excuse me, miss. What? Your shirt's backwards.",
  "transcriptionMethod": "whisper",
  "timestamp": "2026-03-14T12:00:00.000Z",
  "segmentIndex": 0,
  "segmentStart": 0.3,
  "segmentEnd": 1.16,
  "segmentText": "Excuse me, miss.",
  "totalSegments": 3,
  "errMsg": ""
}

Typical Use Cases

• AI agents & LLM pipelines — Speech-to-text ingestion for content understanding
• Content summarization — Get the spoken word from any Instagram video
• Social media analysis — Analyze what creators are saying at scale
• Subtitle generation — Create captions for repurposing content
• Trend & language research — Study speech patterns across video content
• Batch transcription — Process hundreds of videos in one run

Transcription Methods

Auto (Recommended)

Tries native Instagram captions first. If not available, automatically falls back to Whisper AI. Best balance of speed and coverage.

Native

Only uses Instagram's auto-generated captions. Fastest and cheapest, but may not be available for all videos (especially older posts or non-Reels content).

Whisper

Always uses the Whisper AI speech-to-text model. Works for any video with speech. Choose model size based on your accuracy/speed needs:

Model	Size	Speed	Accuracy	Best For
tiny	39 MB	Fastest	Basic	Quick previews
base	74 MB	Fast	Good	Most use cases
small	244 MB	Moderate	Very good	High-accuracy needs

FAQ

Q: Does this work with Instagram Reels? A: Yes! Reels, regular video posts, and IGTV are all supported.

Q: What if the video has no speech? A: The scraper will return an empty transcript with an appropriate message. Music-only or silent videos produce no transcript.

Q: What languages are supported? A: Whisper supports 99+ languages with automatic detection. You can also specify a language code to improve accuracy.

Q: How accurate is the transcription? A: Native captions use Instagram's own speech recognition. Whisper base model provides good accuracy for most content. Use small for better results with accented or complex speech.

Q: Can I process videos in bulk? A: Yes! Provide multiple URLs in the videoUrls array. The scraper processes them sequentially with automatic delays between requests.

Notes

• Transcript accuracy depends on audio clarity and language
• Each transcript segment is a separate dataset row for easy export and filtering
• The scraper uses browser automation with anti-detection measures for reliable access
• Please ensure compliance with Instagram's terms and local regulations