Post-call sales recap video: three-scene AI avatar summary, next step

A post-call recap video is a 30-second Sume avatar clip in three scenes: thanks, what was agreed, and the next step. It costs $5.52 on standard, $7.35 on plus and $16.50 on max. Write the scenes as ordered video_inputs so each beat is explicit, and send them to POST /v1/avatar-1.0/talking-video. The avatar video docs show the scene fields.

Facts only from the call notes

Build the script from your own call notes, not from guesses. Do not state prices, dates or commitments the prospect did not hear on the call, because a recorded clip is easy to forward. Review one preview before the first batch, then reuse the structure.

• Scene ids such as thanks, agreed, next keep the plan readable.
• One avatar and one shared background per video, as Sume currently requires.
• Total planned duration of 4-60 seconds.

Cost per prospect

Thirty seconds, no product image.

Three scenes in one request

Each spoken scene needs a duration and either script or input_text.

curl -X POST https://api.sume.com/v1/avatar-1.0/talking-video \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: recap-call-3318" \
  -d '{
    "avatar_handle": "sales_host",
    "aspect_ratio": "16:9",
    "quality": "standard",
    "video_inputs": [
      {"id": "thanks", "voice": {"type": "text", "script": "Thanks for the call, Jo.", "duration": 4},
       "background": {"type": "prompt", "prompt": "Bright home office"}},
      {"id": "agreed", "voice": {"type": "text", "script": "We agreed to a pilot plan by Friday and three use cases.", "duration": 12},
       "background": {"type": "prompt", "prompt": "Bright home office"}},
      {"id": "next", "voice": {"type": "text", "script": "We meet again next Tuesday. Reply with any changes.", "duration": 10},
       "background": {"type": "prompt", "prompt": "Bright home office"}}
    ]
  }'