Technical Specifications of runway-act-one
| Specification | Details |
|---|---|
| Model ID | runway-act-one |
| Provider | Runway |
| Model type | Character-performance animation / image-and-video-to-video facial performance transfer |
| Core function | Applies a driving performance video to a character reference image or character reference video to animate expressions, lip movement, and facial motion. |
| Base model context | Act-One is presented by Runway as a feature used with Gen-3 Alpha and Gen-3 Alpha Turbo workflows. |
| Primary inputs | Driving performance video plus either a character reference image or a character reference video. |
| Recommended subject type | Human, forward-facing, single-face shots with clear facial features and minimal occlusion. |
| Supported output duration | Up to 30 seconds in the documented Gen-3 Alpha / Turbo Act-One workflow. |
| Output resolutions | 1280×768 and 768×1280 in the documented workflow. |
| Frame rate | 24 fps. |
| Motion control | Motion Intensity setting from 1 to 5, where lower values favor stability and higher values increase expressiveness. |
| Platform availability | Runway documents Act-One availability on web and iOS for the referenced workflow. |
| Access considerations | Runway states this feature is available to Standard-plan users or higher in its product workflow, while API access uses Runway’s developer platform, API keys, versioned headers, and task-based generation flow. |
What is runway-act-one?
runway-act-one is CometAPI’s platform identifier for Runway’s Act-One model/workflow, a character animation system designed to transfer a recorded performance onto a target character. In Runway’s own description, users upload a driving performance and use it to influence a character reference image or video, especially for expressions, mouth movement, and facial acting.
In practice, this makes runway-act-one well suited for talking-character clips, stylized portrait animation, AI avatar shots, and performance-driven facial animation where you want more control than prompt-only video generation. Runway’s guidance emphasizes that the best results come from well-lit, forward-facing, human single-face inputs with the face visible for the full shot.
Although Act-One is described in Runway’s help documentation as part of the Gen-3 Alpha and Turbo creation flow, developers integrating Runway programmatically should understand that Runway’s API is task-based, authenticated by bearer token, and versioned using the X-Runway-Version header.
Main features of runway-act-one
- Performance-driven facial animation: Transfers a source performance onto a target character, allowing expressions and mouth movement to follow the driving video rather than relying only on text prompting.
- Character image or video input: Supports either a still character reference image for more stable outputs or a character reference video for greater motion flexibility.
- Expression and lip-sync influence: Runway explicitly positions Act-One as a way to precisely influence expressions, mouth movements, and related facial behavior.
- Adjustable motion intensity: Includes a Motion Intensity control from 1 to 5 so users can tune outputs toward steadier motion or stronger expressiveness.
- Portrait-friendly production specs: The documented workflow supports both landscape and portrait outputs at 24 fps, with output lengths up to 30 seconds.
- Best-practice driven reliability: Runway provides concrete input guidelines such as shoulders-up framing, forward-facing faces, minimal occlusion, and consistent visibility to improve success rate and quality.
- Task-based API compatibility: For developer workflows, Runway’s API uses authenticated task submission and task retrieval patterns, which fit well into asynchronous production pipelines.
- Reusable asset ingestion options: Runway supports direct HTTPS inputs and ephemeral uploaded assets via
runway://URIs, which can help when automating media-heavy workflows.
How to access and integrate runway-act-one
Step 1: Sign Up for API Key
To call runway-act-one, first create an account on CometAPI and generate an API key from the dashboard. Store it securely as an environment variable so your application can authenticate requests without hardcoding secrets in source files.
Step 2: Send Requests to runway-act-one API
Use CometAPI's Runway-compatible endpoint at POST /runway/pro/act_one.
curl https://api.cometapi.com/runway/pro/act_one \
-H "Content-Type: application/json" \
-H "Authorization: Bearer $COMETAPI_API_KEY" \
-d '{
"prompt": "Your prompt here."
}'
Step 3: Retrieve and Verify Results
The API returns a task object. Poll the task status via POST /runway/feed with the task ID to check when generation is complete, then retrieve the output URL.