Generate AI images and videos with multiple model options, custom trained models, editing, and upscaling via LetzAI API.

• Supports three image models (Nano Banana Pro, Flux2 Max, SeeDream) and three video models (VEO, Kling, Wan) with configurable resolution modes and dimensions
• Includes context-based image editing, skin fixing, and upscaling with strength control (1-3x)
• Custom trained models for persons, objects, and styles accessible via @modelname syntax in prompts
• Asynchronous generation with polling pattern (3-second intervals for images/edits, 2-3 seconds for videos) until status reaches "ready"
• Requires Bearer token authentication and paid LetzAI subscription; credit costs vary by model and resolution