Playground

The playground is a good way to verify that a model meets your needs before writing integration code. Experiment results help you settle on the model call name, prompts, and common parameters.

Chat

1. Open Playground on the website and choose AI Chat.
2. Select a model below the input box.
3. Type a question and send; watch the response content and speed.

Images

1. Choose AI Image.
2. Select a model, then fill in the image description and size.
3. After generation, check whether the picture matches the prompt, then use the same model call name for API integration.

Input Tips

• Describe the subject, action, environment, and style separately.
• For image editing, provide a clear original image and state explicitly which areas to keep and which to change.
• If generation fails, reduce the number of input images or compress the files first, then check whether the model supports the capability.

Video

Video generation is an asynchronous task: submit a generation request, wait for the task to finish, then read the result.

1. Choose AI Video.
2. Select a model and describe the video; when reference images are needed, provide publicly accessible image URLs.
3. Set resolution, aspect ratio, and duration, then submit.

Usage Tips

• Write prompts that clearly cover the subject, actions, camera movement, and environment changes.
• Reference images must be URLs your server can reach.
• Before generating with people or multiple assets, read the Asset Selection Guide.

Available options in the playground depend on the model. For production calls, follow the corresponding API documentation and model capabilities.