Video Transcriber

The Video Transcriber is a powerful tool that extracts transcripts from videos across multiple platforms including Instagram, Twitter/X, Facebook, TikTok, YouTube, and more.
It provides detailed video metadata, timestamped transcripts, and supports multiple languages automatically using OpenAI's Whisper model.

📊 Optimization Summary

• Image Size: Optimized to ~4-5 GB (reduced from 8 GB)
• Build Time: ~5-6 minutes (reduced from 10 minutes)
• NumPy Fix: Pinned to <2.0.0 to avoid compatibility issues

---

✨ Features

• 🔗 Works across multiple platforms:

  • Instagram
  • Twitter/X
  • Facebook
  • TikTok
  • Vimeo
  • Dropbox
  • Google Drive
  • Loom

• 🎧 High-quality, accurate transcription

• 📄 Returns detailed video metadata

• 🕒 Generates timestamped transcript segments

• 🌍 Multi-language support (auto-detection)

• ⚡ Fast and reliable for videos of any length

---

📥 Input

The actor accepts the following input parameters:

{
  "url": "https://www.instagram.com/watch?v=example" // Required: Video URL
}

Parameters

• url (required) → The video URL you want to transcribe

---

📤 Output

The actor outputs a JSON object containing video details and transcript:

{
  "url": "https://www.instagram.com/watch?v=example",
  "title": "Video Title",
  "description": "Video Description",
  "duration": "10:00",
  "viewCount": 1000,
  "uploadDate": "20240911",
  "language": "en",
  "transcript": "Full transcript text...",
  "segments": [
    {
      "start": 0.0,
      "end": 5.0,
      "text": "Transcribed text segment..."
    }
    // ... more segments
  ]
}

---

🚀 Usage

1. Using Apify Console

1. Go to the actor’s page on Apify
2. Enter the video URL
3. Click Run

2. Using Apify API

from apify_client import ApifyClient

client = ApifyClient("YOUR_API_TOKEN")

run_input = {
    "url": "https://www.instagram.com/watch?v=example"
}

run = client.actor("username/video-transcriber").call(run_input=run_input)
results = client.dataset(run["defaultDatasetId"]).list_items().items

---

⚡ Highlights

• Blazing fast transcription 🚀
• Accurate and reliable across platforms 🎯
• Clean JSON output ready for integration 📊
• Fully automated with no manual steps needed 🤖

---

⚠️ Limitations & Troubleshooting

Instagram Rate Limiting

• Issue: Instagram may block access with "rate-limit reached or login required"
• Solution: Provide Instagram cookies in the instagramCookies input field
  1. Install browser extension: "Get cookies.txt LOCALLY"
  2. Visit instagram.com and log in
  3. Export cookies and paste into the input field
  4. This allows consistent access to both public and private content

Other Limitations

• Videos must be accessible (public or you must have authentication)
• Some platforms may require authentication (Instagram specifically)
• Processing time grows with video length (use 'tiny' model for faster results)

Model Selection

• tiny: Fastest, ~5-10x faster than small, good accuracy for most content
• base: Balanced speed/accuracy
• small: Most accurate but slower, use for professional/technical content

---

👨‍💻 Author

Built with ❤️ by Webscrap Team