image-gen

Generate concept art from a text prompt by executing the wu image-gen command with your descriptive text. This produces a single image file path, accelerating visual reference creation and reducing manual artwork iteration.

Yes, you can guide style, lighting, or composition by attaching one or more conditioning images using the --image argument. This directs the text-to-image generation to match your visual references.

The wu CLI supports Gemini, OpenAI, and Nim backends for text-to-image generation. This backend flexibility lets you target specific latency, style, or policy requirements for your generated visuals.

Yes, you need the wu CLI installed and available on your system PATH to generate images from prompts. Once configured, you can execute commands and specify the output path for the resulting image file.

Integrate generated images into design pipelines by specifying an output path with the --output argument. This yields a single image file per invocation, suitable for seeding texture references, material references, or asset previews.

A core limitation is that each invocation outputs only a single image path, and conditioning images are optional inputs. You must manage batching and iteration externally if you need multiple variations for concept visuals.