Session zero welcome clip with house rules on screen, $0.38

To welcome new players before session zero, make a 60-second clip where each house rule appears on screen as it is spoken: generate four rule cards with an image model, record the narration, and join them in one Sume Timeline 1.0 render. At the rate card read on 2026-10-04 the whole thing is about $0.38, and players can rewatch it instead of asking you the same questions in chat.

Putting the rule on screen matters because a spoken list gets forgotten by the third item. A card with four words and a number does not.

Which four rules go on cards?

Pick the ones that cause arguments: table etiquette (phones down in combat), content limits and the pause signal, how dice are rolled, and the schedule. One short quoted line per card, such as "Phones down in combat". Ask for the exact words in the prompt and proofread each output; a typo on a rules card is worse than none.

What does it cost?

Ideogram 4.5 lists $0.03, $0.06 or $0.22 per image by quality, before Sume's margin (Image API docs). Text to speech is $0.0475 per 1,000 characters, and a 60-second read is roughly 900 characters. Timeline is $0.10 per output minute, rounded up (API pricing).

How do I put it together?

The narration is the spine, with its duration_seconds. Each card is a slot; four slots of 15 seconds make 60. All URLs must be on media.sume.com, which generated files already are (Media inputs). Run /v1/timeline-1.0/plan first: it is unbilled and shows the duration and an estimate before you spend.

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: session-zero-001" \
  -d '{
    "audio": { "url": "NARRATION_URL", "duration_seconds": 60 },
    "video": [
      { "source_url": "CARD_1", "start": 0,  "duration": 15 },
      { "source_url": "CARD_2", "start": 15, "duration": 15,
        "transition": { "type": "fade", "duration": 0.5 } },
      { "source_url": "CARD_3", "start": 30, "duration": 15,
        "transition": { "type": "fade", "duration": 0.5 } },
      { "source_url": "CARD_4", "start": 45, "duration": 15,
        "transition": { "type": "fade", "duration": 0.5 } }
    ]
  }'

What should I keep out of the clip?

Keep anything you might change out of the video: dates, the venue, and the link to the group. Put those in the message that carries the clip. Time the card changes to the sentence breaks of your narration, and keep slot durations summing to the spine; coverage may trail it by at most half a second.

If you want music under it, add a soundtrack with duck_db so it sits below the voice; the approach in the session recap post uses the same fields.