Your first image
Still images are the fastest way to see Fawna at work. Open Studio, switch to an image tier, write a descriptive prompt, and get four photoreal renders in seconds.
Pick an image tier
Open the Model drawer and switch to the Image tab. Two tiers to choose from:
Write the prompt
For Illustrate, start your prompt with "A photo of..." if you want photoreal output. Be specific about subject, setting, lighting, lens, and composition. Descriptive adjectives like serene, intimate, sun-warmed steer mood. Concrete nouns like brass, weathered leather, dust motes in beams give the model texture to anchor to.
A photo of an antique brass pocket watch resting on a weathered leather journal, warm amber late-afternoon sunlight streaming across the desk, dust motes visible in the beams, ink quill and wax seal to one side. Shallow depth of field, 50mm macro lens, cinematic composition, intricate texture detail on the engraved brass and aged leather.
Aspect, quality, outputs
- Aspect: Illustrate supports 1:1, 4:5, 3:4, 16:9, 9:16. Pick the one that matches where the image will live.
- Quality: Fast for quick iteration, Standard as the default, Ultra when you're ready for a hero shot (costs more credits).
- Outputs: 1, 2, or 4 per generation. Four gives you variety for the same prompt; each is an independent attempt.
Using references (Compose)
If you pick Fawna Compose, a Characters slot appears in the ref row. Drop an image in to lock that face or object across every generation. Compose preserves facial features, clothing, and distinctive appearance details while obeying the prompt's new scene description.
Illustrate doesn't use reference images. If you attach refs while Illustrate is selected, you'll see a warning in the composer suggesting you switch to Compose.
Where to go next
- Writing a great image prompt for the full playbook.
- Using references for character and style references in depth.