For browser use, sign in and open AI Creation. Select a video model, describe the scene and check the displayed specifications and price before submitting.

​
Use the API with Apifox

Create a key in Token Management and follow Apifox setup. Generation consumes account credit. Check current models and prices; do not repeatedly submit the same generation request. Create a POST request to https://infistar.ai/v1/videos. Set Authorization: Bearer YOUR_API_KEY and Content-Type: application/json. 
{
  "model": "VIDEO_MODEL_ID",
  "prompt": "A wooden boat crosses a misty lake in the morning, cinematic framing.",
  "seconds": "5",
  "size": "1280x720"
}
Replace the model ID and choose duration and size supported by that model. Save the returned id.

​
Poll and download

Create a GET request to https://infistar.ai/v1/videos/TASK_ID using the same authorization header. For queued or in_progress, wait around ten seconds before checking again. For failed, inspect error. Poll with GET; another POST creates another task and may incur another charge. When status is completed, open metadata.url if provided. Otherwise request GET /v1/videos/TASK_ID/content with the same key and save the response using Apifox’s download control. For 401, check the key and the space after Bearer. For insufficient balance, check Wallet. For 404, check the full address and saved task ID. Keep the task ID when asking support about a long-running task.