Sume Format vs pasting the same prompt each time: what changes

A Sume Format saves a production recipe, so your backend calls it by name and sends only the changing inputs, where a pasted prompt sends the whole brief on every request. The recipe is set before your instruction, so you do not resend a system prompt on every call (Format API).

Side by side

One run

The endpoint is POST /v1/formats/{handle}/{slug}/runs and the key needs formats:write. It returns 202 and a receipt with status, result and cancel URLs.

curl -sS -X POST "https://api.sume.com/v1/formats/acme/live-commerce/runs" \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: order-8823-v1" \
  -d '{
    "instruction": "Vertical 9:16 host video. No BGM, no captions.",
    "input": { "product_url": "https://shop.example.com/p/8823" },
    "generation_spend_cap_usd": 120
  }'

How the pieces are ordered

The agent sees the Format pointer and package first, then your instruction, then your input written to a file and read as data. Where the Format and your instruction disagree, the model follows what you asked for.

Not the same as chat

A run is one unattended turn. It does not stop to ask questions, and approvals that a chat-authored recipe asks for are pre-granted within the spend cap. Runs that make video take minutes, with long-form host video usually 15 to 30 minutes, so design for the asynchronous path.