Pika's auto model pick vs Sume Auto and pinned ids

Pika's Sept 17, 2026 post "Welcome to the New Pika" lists 25-plus apps, a choice between automatic and manual model selection, and Seedance generations it calls half the price of other creation platforms. Sume has the same two modes through one API: sume/auto picks for you, and a pinned id such as seedance-2.5 fixes the model.

What Pika announced

The details below come from Pika's post. The price claim is Pika's own statement, not a measured comparison.

Auto on Sume

model: "sume/auto" on POST /v1/videos lets Sume pick the family. The response echoes sume/auto, and Sume does not disclose which family served the request. Resolution is a pure function of the request and the catalog version, so a replay of the same idempotent request prices and routes identically. Auto create controls default to 720p and 8 seconds, with 3-10 second clips at 16:9 or 9:16.

The trade-off is that you cannot rely on a visible trait of the output to know which model ran.

Pinned ids for fixed cost

Pin an id when you need repeatable look and a known price. seedance-2.5 accepts 4-30 seconds at 480p, 720p and 1080p, and Sume bills provider list times 1.25 on every model. Read pricing_skus for the id from GET /v1/videos/models and multiply by your clip length.

Both calls use the same endpoint; only the model value changes.

# let Sume pick
curl -X POST https://api.sume.com/v1/videos \
  -H "Authorization: Bearer $SUME_API_KEY" -H "Content-Type: application/json" \
  -H "Idempotency-Key: auto-001" \
  -d '{"model":"sume/auto","prompt":"A sneaker on a studio plinth","aspect_ratio":"9:16","duration":5}'

# pin the model
curl -X POST https://api.sume.com/v1/videos \
  -H "Authorization: Bearer $SUME_API_KEY" -H "Content-Type: application/json" \
  -H "Idempotency-Key: pinned-001" \
  -d '{"model":"seedance-2.5","prompt":"A sneaker on a studio plinth","aspect_ratio":"9:16","duration":5}'