Image Generation

Generate images from text prompts using models like Flux and SDXL:

melious image "a sunset over the Swiss Alps, oil painting style"

Images are saved to the current directory as melious-{timestamp}.{ext} and — on supported terminals — displayed inline.

Options

Flag	Type	Default	Description
`--model`	string	`flux-1-schnell`	Image generation model
`--size`	string	`1024x1024`	Image dimensions (e.g., `1792x1024`)
`--quality`	string	`standard`	Quality: `standard` or `hd`
`--style`	string	`vivid`	Style: `vivid` or `natural`
`-n`, `--count`	int	`1`	Number of images to generate
`-o`, `--output`	string	`.`	Output directory (created if needed)
`--seed`	int	—	Random seed for reproducibility
`--steps`	int	—	Inference steps (model-dependent)
`--guidance`	float	—	Guidance scale (model-dependent)
`--negative`	string	—	Negative prompt

Generate multiple images:

melious image -n 4 "cyberpunk cityscape at night"

Save to a specific directory:

melious image -o ./renders "abstract geometric patterns"

Use a specific model with higher quality:

melious image --model flux-1-dev --quality hd "portrait of a cat reading a book"

Reproduce a specific result:

melious image --seed 42 "minimalist logo design"

Fine-tune generation:

melious image --steps 50 --guidance 7.5 --negative "blurry, low quality" "a serene mountain lake"

The CLI attempts to render images inline in your terminal if supported:

If your terminal doesn't support inline images, the CLI saves the file and prints the path.

Use --json to get metadata without the image data:

melious image --json "a red circle"

This returns model info, dimensions, and file paths but omits the base64 image data.