Product hero from 3 angle photos with Flux 2 Pro: 100 for $3.75

Answer

Flux 2 Pro on Sume accepts up to 10 reference images at a flat $0.0375 per image, so a hero shot built from three angle photos of a product costs $0.0375, 100 heroes cost $3.75 and 1,000 cost $37.50. The three references do not change the price, because the row is billed per output image.

Put the best angle first and say in the prompt what stays fixed: the shape, the label and the colorway. Then describe the new setting, light and crop.

Request

curl -X POST https://api.sume.com/v1/images \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"black-forest-labs/flux.2-pro","prompt":"hero shot of the product from the references on a wet slate counter, rim light, label unchanged","aspect_ratio":"4:5","n":2,"input_references":[{"type":"image_url","image_url":{"url":"https://example.com/front.jpg"}},{"type":"image_url","image_url":{"url":"https://example.com/side.jpg"}},{"type":"image_url","image_url":{"url":"https://example.com/top.jpg"}}]}'

What to check on each output

• Label text and logo: compare against the packshot. An image model can alter small print.
• Proportions: the product should match the reference angles, not a generic version of the product type.
• Scene logic: contact shadows, reflections and the surface under the product.
• Keep the original packshot as the listing's source of truth. Treat the generated hero as marketing art.

Flux 2 Pro lists 13 aspect ratios and n up to 4. It does not accept quality, resolution, mask_url or background, and a request that sets one returns a 400. Failed generations are not billed. Docs: Sume Image API.

Before a large run

Prices and descriptors change when the catalog changes, so confirm them before you spend. Call GET /v1/images/models/{id}/endpoints for the row you plan to use and read its pricing line and supported_parameters; both come back in one response.

Then run a pilot of three to five images and read usage.cost on each response. Multiply by your planned count for a forecast you can trust. Completed generations are billed in full and failed or cancelled ones are not, so a pilot that errors costs nothing.

For big batches, use mode: "async" or mode: "webhook" with a public HTTPS webhook_url, so no request waits on the 30-second sync limit. Poll GET /v1/jobs/{id}/status and fetch GET /v1/jobs/{id}/result when the job completes.