Technical Specifications of runway-video
| Specification | Details |
|---|---|
| Model ID | runway-video |
| Provider | Runway |
| Model type | AI video generation |
| Primary modalities | Text-to-video and image-to-video, depending on the underlying Runway model selected |
| Inputs | Text prompts, reference images, and in some cases uploaded media assets |
| Outputs | Generated video clips |
| API pattern | Asynchronous task-based generation workflow |
| Authentication | Bearer API key authentication |
| Required version header | X-Runway-Version: 2024-11-06 |
| Base API host | api.dev.runwayml.com |
| Official SDKs | Node.js and Python SDKs are available |
| File handling | Supports temporary uploads that return runwayUri values for generation requests |
| Pricing model | Credit-based billing; available Runway video models are billed per second of generated output |
| Example underlying video models | gen4.5, gen4_turbo, gen4_aleph, veo3, veo3.1, veo3.1_fast, gen3a_turbo |
What is runway-video?
runway-video is CometAPI’s platform identifier for accessing Runway’s video-generation capabilities through a unified API. Based on Runway’s official API documentation, Runway offers multiple video models for generating clips from text prompts or from images, with task-based endpoints for starting generations and later retrieving results.
In practice, runway-video is best understood as a gateway to Runway-powered video creation workflows rather than a single standalone foundation model name. Runway’s current API catalog includes several video models with different input modes and pricing tiers, such as gen4.5 for text-or-image-to-video, gen4_turbo for image-to-video, and additional options like gen4_aleph, veo3, veo3.1, veo3.1_fast, and gen3a_turbo.
This makes runway-video suitable for applications such as creative prototyping, marketing content generation, social media clip production, concept visualization, and product experiences that need automated short-form video synthesis. Because the underlying Runway API uses asynchronous jobs, integrations typically submit a generation request first and then poll or retrieve task output once processing finishes.
Main features of runway-video
- Text-to-video generation: Runway supports video creation from natural-language prompts on supported models, enabling developers to generate clips directly from descriptive text.
- Image-to-video workflows: Several Runway video models accept image inputs, making
runway-videouseful for animating still images or building reference-driven video generation pipelines. - Multiple underlying model options: The Runway API exposes several video models with different capabilities and cost profiles, giving developers flexibility to optimize for quality, speed, or budget.
- Asynchronous task processing: Generation requests create tasks instead of returning final media immediately, which is well suited for production systems that need queueing, retries, and status checks.
- Temporary asset uploads: Runway provides upload endpoints that return
runwayUrireferences, making it easier to pass images, video, or audio assets into later generation requests. - Official SDK support: Runway offers Node.js and Python SDKs, which can simplify authentication, request construction, and output retrieval.
- Versioned API integration: Requests require an explicit
X-Runway-Versionheader, which helps developers build against a known API contract. - Credit-based pricing: Runway bills video generation by credits per second, with pricing varying by the selected underlying model.
How to access and integrate runway-video
Step 1: Sign Up for API Key
To access runway-video, first create an account and generate an API key through your CometAPI dashboard. After signing in, copy your API key and store it securely in an environment variable or secrets manager for production use.
Step 2: Send Requests to runway-video API
Use your API key to authenticate requests to the CometAPI endpoint for runway-video. Pass your prompt and any supported input parameters in the request body, and specify the model as runway-video.
curl -X POST "https://api.cometapi.com/v1/responses" \
-H "Authorization: Bearer $COMETAPI_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "runway-video",
"input": "Create a cinematic video of a futuristic city at sunrise with slow camera movement."
}'
Step 3: Retrieve and Verify Results
After submission, inspect the API response for the generated output, task identifier, or result payload returned for runway-video. If the generation runs asynchronously, poll the relevant CometAPI endpoint until processing completes, then validate the returned video URL, metadata, and content before using it in your application.