AI voiceover for short videos: 9:16 with voice and music

An AI voiceover for a short video is a spoken track made from your script, laid under vertical clips. On Sume: generate the voice with POST /v1/tts-1.0/generate, then assemble the clips on that voice with POST /v1/timeline-1.0/render, whose default output is a 1080x1920 MP4. Add a music bed and burned-in captions as optional steps.

Facts are from the Timeline 1.0, Video captions and Music Router docs and the Sume API reference, read 2026-09-29.

What are the steps?

How do I keep the music under the voice?

Use the render's soundtrack field: it takes a url, gain_db, loop, fade_out_seconds up to 10 and duck_db 0-20, and ducking needs a real voice spine, not silence.

curl -X POST https://api.sume.com/v1/timeline-1.0/render \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: short-001" \
  -d '{
    "audio": { "url": "https://media.sume.com/artifacts/artf_demo/voice.wav", "duration_seconds": 24 },
    "video": [
      { "source_url": "https://media.sume.com/artifacts/artf_demo/a.mp4", "start": 0, "duration": 12 },
      { "source_url": "https://media.sume.com/artifacts/artf_demo/b.mp4", "start": 12, "duration": 12 }
    ],
    "soundtrack": { "url": "https://media.sume.com/artifacts/artf_demo/bed.mp3", "gain_db": -18, "duck_db": 8, "loop": true }
  }'

How long can the short be?

The render's audio.duration_seconds runs 1 to 1,800, well beyond any short-form cap, so your platform's limit decides. Check the platform's own current rules for length before you export.