Runway Video API Configuration
Runway is a video generation platform. It is usually not the Base URL to fill in Chatbox. Instead, create an API Key on the Runway platform, and then use the official SDK or HTTP request to create video tasks.
Official information Downloads and documentation are subject to the official site.
What scenarios is it suitable for
Text-to-VideoImage-to-VideoAsynchronous TasksOfficial SDK
First understand the video task workflow
- Create an API Key and make sure your account has video API permissions and balance.
- Select the task type: text-to-video, image-to-video, or other video tasks currently supported by the official documentation.
- You will not get the video file immediately after submitting a task; usually, you will first receive a task object or task ID.
- Wait for the task to complete, then read the output address or download the results.
- In case of failure, first check the task status and error details, do not just check the Key.
Prepare parameters
| Configuration item | Filling Method |
|---|---|
| API Key | Runway backend creation; the official SDK reads <code>RUNWAYML_API_SECRET</code> by default |
| Task | text-to-video / image-to-video |
| Prompt | Video content description |
| Ratio / Size | Landscape, portrait, or specified size |
| Duration | Video Duration |
| Status | Poll the task until completion |
| Setting | Description | Tips for beginners |
|---|---|---|
| API Key | Request credentials for the Runway API | Only save in the local machine or server environment variables, using the variable name RUNWAYML_API_SECRET |
| Prompt | Video content description | Write shorter first, then refine after confirming the style |
| Image | The starting image of the image-to-video | Upload clear images with a clear subject |
| Duration | Video seconds | First test the cost and effect with a shorter duration |
| Status | Task Status | Video generation is asynchronous and requires waiting for completion. |
SDK Example
Runway environment variable name
RUNWAYML_API_SECRET=YOUR_RUNWAY_API_KEYRunway Node.js Task Example
import RunwayML, { TaskFailedError } from "@runwayml/sdk"; const client = new RunwayML(); try { const task = await client.imageToVideo.create({ model: "gen4.5", promptText: "A product introduction video on a white background with a slow camera push-in", ratio: "1280:720", duration: 5 }).waitForTaskOutput(); console.log(task.output[0]); } catch (error) { if (error instanceof TaskFailedError) { console.error(error.taskDetails); } }The fields of the video platform will be updated with the model version. The page example is used to understand the configuration structure, and the actual fields should be based on the current version of the official documentation.
| Phenomenon | Priority check | Handle |
|---|---|---|
| The video is not returned immediately after submission | Is the task still in the queue or being generated? | Video generation is an asynchronous task; wait until the status is complete before retrieving the result. |
| Mission failed | Model name, scale, duration, whether the input image meets the current model requirements | Adjust the parameters according to the returned error details, and do not change too many items at once. |
| The charges are higher than expected | Duration, resolution, model version | First test with short duration and low-cost parameters, then generate in bulk |