Subscut video clipping API documentation

This endpoint is built specifically for spoken long-form video. You send one public video URL and receive multiple rendered clips back in a single response.

Each clip includes a title, a quality score, clip boundaries, and the final rendered video_url.

You can authenticate with either header format:

• x-api-key: subscut_your_api_key
• Authorization: Bearer subscut_your_api_key

Create keys from Dashboard API Access.

Public request shape:

{
  "video_url": "https://example.com/podcast.mp4",
  "max_clips": 5,
  "clip_duration": {
    "min": 20,
    "max": 60
  },
  "captions": true,
  "caption_position": "center",
  "style": "viral"
}

{
  "clips": [
    {
      "start": 124.2,
      "end": 156.8,
      "score": 0.93,
      "title": "Why Most Founders Get This Wrong",
      "video_url": "https://your-public-s3-base.example/generated/api-rendered/.../clip-1.mp4"
    },
    {
      "start": 381.1,
      "end": 414.6,
      "score": 0.88,
      "title": "Nobody Talks About This Tradeoff",
      "video_url": "https://your-public-s3-base.example/generated/api-rendered/.../clip-2.mp4"
    }
  ]
}

curl -X POST https://subscut.com/podcast-to-clips \
  -H "Content-Type: application/json" \
  -H "x-api-key: subscut_your_api_key" \
  -d '{
    "video_url": "https://example.com/podcast.mp4",
    "max_clips": 5,
    "captions": true,
    "caption_position": "center",
    "style": "viral"
  }'

If you want another agent or LLM tool runtime to call the API reliably, give it this compact contract:

Endpoint: POST /podcast-to-clips
Auth: x-api-key header or Bearer API key
Goal: turn a long spoken video into multiple short clips

Required:
- video_url: public http/https URL to a podcast/interview/talking-head video

Optional:
- max_clips: integer 1-10
- clip_duration.min: integer seconds
- clip_duration.max: integer seconds
- captions: boolean
- caption_position: "center" | "bottom"
- style: "viral" | "leon" | "hormozi" | "clean" | "minimal"

Output:
- clips[] with title, score, start, end, and rendered video_url

Use when:
- the user wants to repurpose a podcast into Shorts, Reels, or TikTok clips

Do not use when:
- the source is already short-form
- the content is mostly music-only or non-speech

• video_url must be publicly reachable by the server.
• The endpoint is optimized for speech-heavy content like podcasts, interviews, and talking-head videos.
• If the source is already short-form or mostly non-speech, the output quality will be weaker.