Video Transcription API

Transcription captures what was said. VidContext captures what was said, what was shown, what appeared on screen, what brands were visible, and what the audio environment sounded like. One API call gives you the full picture — not just the script.

Why transcription alone is not enough

A product demo where the presenter says "click here" is meaningless without knowing what "here" refers to. A cooking video where someone says "add this" loses context without the visual. An ad where the brand name only appears as a logo on screen — and is never spoken — is invisible to transcription.

VidContext bridges this gap. The transcript is paired with timestamped visual scene descriptions, so you know exactly what was on screen when each word was spoken. On-screen text is extracted separately, catching everything from lower-third titles to URL watermarks. The result is a text output that preserves the full meaning of the video, not just the audio track.

Everything you get beyond the transcript

Transcript with full video context

The response pairs every transcript segment with what was happening visually, so you never lose context.

curl -X POST https://api.vidcontext.com/v1/analyze \
  -H "X-API-Key: vc_your_key" \
  -F "source=https://example.com/webinar.mp4" \
  -F "output_format=context"

{
  "transcript": [
    {
      "time": "0:00-0:18",
      "text": "Let me show you how the dashboard works. Click on analytics..."
    },
    {
      "time": "0:18-0:35",
      "text": "You can see the traffic graph updating in real time..."
    }
  ],
  "visual_scenes": [
    {
      "time": "0:00-0:10",
      "description": "Presenter visible in webcam overlay, screen shows app login page"
    },
    {
      "time": "0:10-0:30",
      "description": "Screen recording: dashboard with sidebar, cursor clicks Analytics tab, line chart appears"
    }
  ],
  "on_screen_text": ["Analytics Dashboard", "Last 30 Days", "2,847 visits"],
  "brands_detected": ["Google Analytics"],
  "audio": { "music": "light ambient", "effects": ["mouse click"] }
}

When context-aware transcription matters

• Tutorial and training documentation — Convert video tutorials into step-by-step guides where each instruction is paired with what was on screen, not just what was said.
• Meeting and webinar archives — Capture not just the discussion, but every slide, chart, and screen share that was presented during the meeting.
• Video SEO and indexing — Make video content searchable by every element: spoken words, displayed text, visual content, and brand references.

Frequently asked questions