Text to image, built for quick drafts
Describe subject, lighting, lens, and mood in plain language. MediaMuse routes your prompt to fast Z-Image or higher-quality GPT Image 2, keeps every result in Activity and Assets, and shows the credit estimate before you generate.
Prompt directions for better results
These reusable briefs explain what to describe, what the workflow returns, and where a first attempt can go wrong.
Product and commerce hero
Create a clean product image with material detail, controlled light, and room for a message.
- Make clear
- Name the product, material, surface, light direction, and where copy should sit.
- Avoid
- Avoid asking for invented brand marks or dense packaging copy.
Realistic portrait or editorial photo
Build a believable portrait with natural texture, a specific lens feel, and intentional imperfections.
- Make clear
- Describe the subject, expression, wardrobe, light, lens, and setting rather than only naming a style.
- Avoid
- Avoid vague requests such as “make it realistic” without a subject or light plan.
Social media poster
Make one message legible in a vertical layout with a clear focal point and usable margins.
- Make clear
- Provide the topic, exact short headline, audience, aspect, palette, and the one action the viewer should notice.
- Avoid
- Do not rely on the model for long, exact copy; reserve space and add final text in a design tool.
Explainer infographic
Turn a small set of facts into a readable visual sequence without pretending to know missing data.
- Make clear
- Bring the verified facts, module order, audience, and desired output size; use the image for structure and polish.
- Avoid
- Avoid treating generated numbers or tiny labels as final facts without human review.
Character design sheet
Explore a consistent character across views, poses, and costume details on one reference sheet.
- Make clear
- Lock the character identity, outfit, age, proportions, views, and background before adding decorative details.
- Avoid
- Avoid changing several identity-defining details in one revision if consistency matters.
Document and editorial layout
Set up a readable page with hierarchy, spacing, and visual placeholders for a guide or editorial asset.
- Make clear
- State the page size, reading order, section hierarchy, whitespace needs, and which facts are already verified.
- Avoid
- Treat generated body copy as layout guidance, not a finished document or source of truth.
Live catalog
Models and tools you can use now
Compare the live image recipes, their supported formats, and the workflow each entry opens. Everything here is wired to a real product page.
Live models
What you can do
Two free starter runs
Verified new accounts receive 2 Z-Image generations — enough to test prompts before buying credits.
Model choice in one panel
Switch between Z-Image for speed and GPT Image 2 when you need sharper detail, text, or 2K output.
Batch up to four images
Generate 1–4 stills per prompt with style presets and prompt enhancement — each output is a traceable task.