Generate video using Kling AI (V3) via n8n

Step-by-step

1. Navigate to the KIE AI dashboard documentation and select the Kling V3 API.
2. Copy the cURL command for the POST request to create a video.
3. In n8n, use the 'Import cURL' feature within an HTTP Request node to paste the command.
4. Insert your KIE AI API token into the header.
5. Modify the prompt and model parameters (e.g., change model to 'v3_fast') in the JSON body.
6. Execute the node to receive a Task ID.
7. Set up a GET request node using the Task ID to poll for the video status.
8. Create an 'If' node to check if the 'complete_time' field is populated.
9. Add a 'Wait' node (e.g., 30 seconds) to loop back to the status check if the video is not yet ready.

Options

• Input an image for image-to-video generation
• Set a callback URL to trigger a webhook upon completion instead of polling
• Toggle watermarking on or off
• Adjust aspect ratio
• Switch between standard and fast models

Watch out for

• The callback URL method is more complex as it requires setting up a separate webhook within n8n.

Tips

• Use the KIE AI dashboard to manage just one token for multiple different AI models.
• Set the polling loop to wait 30 seconds rather than 3 seconds to avoid unnecessary API calls, as video generation typically takes a minute or more.
• Check the 'complete_time' field to determine if the generation process is finished.

Highlights

“This is the only token we have to manage. We don't have to deal with all the different API keys that we get from all the different models. We just deal with one.”