Getting started

1. Open a tool or Agent

Use AI Tools when you already know the output type. Use Agent when the request needs planning or several connected steps.

Some actions require an account. Available sign-in methods depend on the authentication providers currently configured for the Service.

2. Choose method, type, and model

Start with the workflow that matches your source material:

• Text to image/video: begin with a written description.
• Image to image/video: upload a reference and describe the change or motion.
• Video to video: upload footage and describe the desired transformation.
• Music: provide a concept, style, lyrics, or instrumental direction using the fields shown.

After you choose a model, review the fields again. Supported references, duration, aspect ratio, resolution, negative prompt, seed, and other controls vary by model.

3. Review cost and submit

The panel shows the current credit estimate or requirement before submission. Cost can depend on model, resolution, duration, output count, and other parameters. Do not rely on an old screenshot or a fixed table; use the value currently shown in the product.

4. Follow task status

Tasks can be queued or processed by an external provider. You may leave the page and use the activity or media area to review status later. Agent runs show a plan with separate steps and may pause, retry a temporary failure, or request your review.

Do not resubmit a slow task repeatedly, because a provider may already have accepted it.

5. Review and save results

Open the completed task, compare results, and download important files. Storage duration is not guaranteed to be permanent. Before publishing, review accuracy, continuity, logos, text, consent, and third-party rights.

For usage conditions, read the AI Output & Content License. For failed tasks and billing issues, read the Refund Policy and contact Support with the task or order identifier.