Social Transcript Extractor — SupaData API

Extract transcripts from video content across all major social platforms via SupaData. Primary use: competitor ad research and content repurposing.

Supported Platforms

• YouTube
• Instagram (Reels, IGTV, Stories)
• TikTok
• Twitter/X (video tweets)
• Facebook (video posts)

Auth & Setup

API key: D:\Ecosystem\secrets\MASTER_API_KEYS.env → SUPADATA_API_KEY

Dashboard: https://dash.supadata.ai

Pricing: Free tier = 100 credits/mo | Pro = $17/mo for 3,000 credits

Basic Extraction

curl -X GET "https://api.supadata.ai/v1/transcript?url=VIDEO_URL&text=true" \
  -H "x-api-key: $SUPADATA_API_KEY"

| Parameter | Required | Description | |-----------|----------|-------------| | url | Yes | Full video URL | | text | No | Return plain text instead of timestamped JSON | | lang | No | Language code (default: auto-detect) |

Python Extraction

import os
import requests

def extract_transcript(video_url):
    response = requests.get(
        "https://api.supadata.ai/v1/transcript",
        params={"url": video_url, "text": "true"},
        headers={"x-api-key": os.environ["SUPADATA_API_KEY"]}
    )
    data = response.json()
    return data.get("content", "")

# Competitor TikTok
transcript = extract_transcript("https://www.tiktok.com/@competitor/video/123456")
print(transcript)

Response Format

{
  "content": "Full transcript text...",
  "lang": "en",
  "availableLangs": ["en", "es"],
  "wordCount": 312
}

Workflow: Competitor Ad Research

1. Find competitor video via TikTok/Instagram/YouTube search
2. Extract transcript with this skill
3. Analyze hook structure, pacing, CTAs
4. Feed into [[scriptwriter]] as reference for new scripts
5. Identify winning angles for [[ai-video-pipeline]]

Workflow: Content Repurposing

1. Extract transcript from Mike's or client's existing video
2. Clean up with Claude
3. Use as blog post body or email newsletter
4. Add to PAA content engine for keyword optimization

Proxy Rule

Instagram and TikTok extraction must use proxies — do not pull from home IP.

Read proxy config from D:\Ecosystem\secrets\MASTER_API_KEYS.env (PROXY_1).

proxies = {
    "http": f"http://{user}:{password}@{host}:{port}",
    "https": f"http://{user}:{password}@{host}:{port}"
}
response = requests.get(url, params=params, headers=headers, proxies=proxies)

Related Topics

• [[assemblyai-transcription]] — transcribe audio files (not social URLs)
• [[ytdlp-agent]] — download YouTube videos and subtitles directly
• [[yt-transcript-pipeline]] — automated YouTube transcript vault

#video-sop #transcript #social-media #youtube #tiktok #instagram #competitor-research