Coming Soon POST /v1/media 10 credits per minute of audio

Any Video In.
Clean Transcript Out.

Point CrawlHQ at a YouTube video, podcast, Instagram Reel, or any public media URL. Get back a clean, timestamped transcript — speaker-attributed, chunked for RAG, structured JSON.

terminal 
curl -X POST https://api.crawlhq.dev/v1/media \
  -H "X-API-Key: $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "url": "https://youtube.com/watch?v=...",
    "output": "transcript",
    "chunk_for_rag": true
  }'
response 
{
  "transcript": [{
    "start": 0,
    "end": 12.4,
    "speaker": "Speaker 1",
    "text": "Welcome to the show..."
  }],
  "duration_seconds": 3482,
  "credits_used": 580
}
200 OK 10 credits per minute of audio

What makes it production-grade

Every module is built for pipelines that run without you watching.

🎬

Universal URL Support

YouTube, Instagram Reels, Twitter/X videos, Spotify podcasts, Vimeo, Loom, and direct MP3/MP4 URLs. One endpoint handles all public media sources.

🎙️

Speaker Diarization

Automatically identify and label distinct speakers in multi-person recordings. Podcasts, interviews, panel discussions — each speaker's words attributed separately.

⏱️

Timestamped Segments

Every transcript segment includes precise start/end timestamps in seconds. Build video search, chapter navigation, or jump-to-quote features on top.

🧩

RAG-Ready Chunks

Set chunk_for_rag: true to receive semantically chunked transcript segments with overlap, ready to embed and index. No additional preprocessing required.

🌐

Multilingual Transcription

Whisper-powered transcription supports 50+ languages. Auto-detect language or specify explicitly. Output in original language or auto-translated to English.

📄

Structured Metadata

Response includes video title, channel, duration, upload date, view count, and description — sourced from the platform alongside the transcript.

Use Cases

What teams build with media

Podcast Intelligence Platform

Transcribe thousands of podcast episodes automatically. Build a searchable archive of industry conversations, expert opinions, and market signals.

Political Speech Analysis

Transcribe candidate speeches, press conferences, and campaign videos. Extract quotes, detect position changes, and build a searchable political archive.

Competitive Intelligence from Video

Transcribe competitor webinars, product demos, and conference talks. Extract announcements, feature roadmaps, and pricing signals automatically.

RAG Knowledge Base from Media

Build AI assistants that can answer questions from your video library. Transcribe, chunk, embed, and query — the full pipeline from a single API.

News & Media Monitoring

Transcribe broadcast news segments, press briefings, and analyst calls. Build structured archives of spoken content alongside web-scraped text.

Education & Training Content

Transcribe lecture recordings, training videos, and webinar archives. Create searchable, subtitled content from raw video at scale.

Frequently asked questions

When is media launching?
Targeting Q4 2026. Join the waitlist for early access and free transcription credits.
What's the pricing model for long videos?
10 credits per minute of audio, rounded up to the nearest minute. A 1-hour podcast = 600 credits. On Starter pricing (₹0.40/credit), that's ₹240 per hour of audio.
Does it work with private or paywalled content?
No. The media module only works with publicly accessible URLs. Authentication support (e.g., YouTube member-only content) is not on the roadmap.
How accurate is the transcription?
We use OpenAI Whisper large-v3, which achieves 3-5% word error rate on clean audio. Noisy recordings, heavy accents, or technical jargon may have higher error rates. You'll receive a confidence_score per segment.
Can I get just the transcript without speaker labels?
Yes. Speaker diarization is opt-in. Set diarize: false (default) for a faster, cheaper transcription without speaker attribution. Diarized transcription costs 2x credits.
What's the maximum video duration?
No hard limit, but videos over 3 hours are processed asynchronously. You'll get a job_id back immediately and a webhook fires when the transcript is ready.

Be first when media ships

Join the waitlist. We'll give you early access and free credits when it launches.

Join Waitlist →