Image edits

The endpoint accepts JSON and multipart form data. For the Wan 2.7 example, supply 1–9 images as uploads or supported references, plus a 1–5,000 character prompt. With modelrush/wan2.7-image or modelrush/wan2.7-image-pro, each source image must be at least 240×240 pixels.

bash
curl https://api.modelrush.ai/v1/images/edits \
  -H "Authorization: Bearer $MODELRUSH_API_KEY" \
  -F "model=modelrush/wan2.7-image-pro" \
  -F "prompt=Keep the character, replace the background with a clean developer studio" \
  -F "image=@source.png"

For Wan 2.7, use n for 1–4 outputs, size for the output dimensions, and watermark when required. Preserve the returned request identifiers.

Set the image count only with top-level n (default 1); it is the sole quantity used for the quote and the request. This endpoint does not forward provider parameters. If a JSON body includes parameters.n, it must be the same integer as n; otherwise the request returns HTTP 400 invalid_n before anything is charged.

Open a model page for its request parameters and current prices. The four Wan 3 landing pages include a developer API section below the Playground. Read GET /v1/models for current schemas and key-specific access; GET /api/model-pricing returns public prices without authentication; with a valid Bearer API key, it returns the prices applicable to that account.

Qwen Image Edit Spicy, Face Swap Spicy and Head Swap Spicy use POST /v1/images/edits, not the image generation endpoint. Their result is asynchronous (HTTP 202): save poll_url and read outputs after status succeeded. Face and head swaps additionally require the account's portrait-rights attestation and per-request person_rights_confirmed; use each model's schema for image ordering and required fields.