skill-perstudio_skill

Overview

This skill generates high-quality images, videos, and audio from natural language prompts. I route user requests to the appropriate model automatically and support transforms like image-to-image, inpainting, upscaling, and animation. It’s optimized for product shots, portraits, stickers, and cinematic video generation.

How this skill works

Send a plain-language intent describing the desired output and optional assets (uploaded images or masks). The skill selects the best generation mode, runs the job synchronously or asynchronously, and returns asset IDs and results. Generation consumes tokens; use the balance action to check remaining credits.

When to use it

• Create photos, product shots, portraits, stickers, or stylized images from text prompts.
• Transform or restyle an existing image with image-to-image, inpainting, or style transfer.
• Upscale images (4x) for higher-resolution outputs.
• Generate short videos or animate still photos with subtle motion.
• Produce voice or text-to-speech outputs using a voice reference.

Best practices

• Be specific in your prompt: include style, lighting, composition, and mood for predictable results.
• Upload a reference image or mask when doing image-to-image, inpainting, or animation for precise control.
• Use auto_upscale: true for automatic resolution enhancement when you need higher detail.
• Check your token balance before large jobs; video generation has a higher token cost.
• Start with smaller test prompts to validate style before spending tokens on full-resolution outputs.

Example use cases

• Generate a product photography shot with studio lighting for an e-commerce listing.
• Create a portrait or avatar in a specific art style by describing lighting and expression.
• Transform a photo into a watercolor painting or vintage film look using image-to-image.
• Animate a still portrait with subtle hair or clothing motion for social media.
• Generate a short cinematic clip like 'a cat playing piano' for concept visualization.