Runway API adds Seedream 5 Pro Layers, then bills it per returned image (November 10 and 11, 2026)
New seedream5_pro_layers model splits an image into layers. Since Nov 11 it costs 3 or 5 credits per returned image, not a flat 18.
Your keys
0 of 1 ready
Create each key on its own page, paste it below, then copy the finished lines into your project's .env file. What you paste stays in this tab. It is never sent to Wirestack or saved.
- Runway API secretRUNWAYML_API_SECRETGet key
Create or pick a project, then open the API Keys tab. Add credits in the Billing tab first
What changed
On November 10, 2026 Runway added Seedream 5 Pro Layers (seedream5_pro_layers) to the API. It splits one image into editable layers. You send one PNG or JPEG source image to the text to image endpoint, with an optional prompt and a resolution of auto, 1k, 1.5k or 2k. A successful task returns a base image and transparent PNG layers. When provided, the response also includes layer names, stacking order and bounds.
On November 11, 2026 the billing changed. At launch a decomposition cost a flat 18 credits at any resolution. It is now billed per returned image:
- 3 credits per image at 1K or 1.5K
- 5 credits per image at 2K or auto
The count includes the base image and each layer. A decomposition returns at most 17 images (a base image plus 16 layers). It reserves 51 credits at 1K or 1.5K and 85 credits at 2K or auto, then settles to the number of images actually returned. A task needs enough credits for the full reservation to start. Runway's example: a 2K decomposition that returns a base image and 8 layers costs 45 credits. Credits cost $0.01 each.
Source image rules: exactly one image, PNG or JPEG only (no WebP), at least 262,144 pixels and at most 36 megapixels in total, aspect ratio between 1:16 and 16:1, neither side over 8,000 pixels, 30 MB or smaller. The usual lower limits still apply: 16 MB for URLs and 5 MB for data URIs.
Prerequisites
- A Runway API key in
RUNWAYML_API_SECRET - Node:
npm install --save @runwayml/sdk(4.25.0 added the Seedream Layers output controls) - Python:
pip install runwayml - Enough credits for the full reservation: 51 at 1K or 1.5K, 85 at 2K or auto
Steps
- Pick one PNG or JPEG source image that meets the input rules above.
- Start a task on the text to image endpoint,
POST https://api.dev.runwayml.com/v1/text_to_image, with the headerX-Runway-Version: 2024-11-06, the modelseedream5_pro_layers, your source image inreferenceImages(for example ), an optional and an optional of (the default), , or .
