Choose inputs, a model and a mode for your task.
- 1
Start with the result
Use model search and task filters. A text-to-image model needs a prompt; image-to-video models also need a source image. Model settings show available inputs and options.
- 2
Review price and access
Prices depend on the model and selected settings. Premium models and quality options explain their access requirements before generation.
- 3
Describe a broader task
Agent mode can plan a task and select tools. Describe your goal and constraints clearly, and review any requested approval before a paid action. For a single generation, select a specific model.