Design Prompt
Generator.
Free tool for Midjourney, DALL·E 3, Stable Diffusion, Flux, Runway, and Figma AI. Describe your scene — get a tool-specific prompt with style, lighting, and negative prompt already dialed in.
Three variations per brief: Balanced, Vivid & painterly, and Minimalist editorial.
Describe what you want to create
3 prompt variations
Tuned for MidjourneyCinematic, Dramatic and moody, Golden-hour natural light, 85mm portrait lens, f/1.4 --ar 16:9 --stylize 200 --v 6.1
Cinematic with painterly flourish, Dramatic and moody, Golden-hour natural light, 85mm portrait lens, f/1.4, volumetric light; fine-grain film emulation --ar 16:9 --stylize 200 --v 6.1
Editorial cinematic, Minimalist, restrained, high-negative-space, Golden-hour natural light, 85mm portrait lens, f/1.4, negative space 60% --ar 16:9 --stylize 200 --v 6.1
Prompt craft cheat-sheet
The 5 ingredients of every great image prompt.
A concrete noun with a verb. 'Astronaut walking' beats 'space scene.'
Cinematic, editorial, 3D render, watercolor — pick one, not three.
Golden-hour, studio softbox, neon backlight. Directs the mood.
Lens + aperture. 85mm f/1.4 tells the model to blur the background.
Watermarks, text, extra fingers. Keeps weird artifacts out.
Related guides
Go deeper with tool-specific guides.
FAQ
Questions about image prompts.
Which image generators does this support?+
Midjourney v6/v7, DALL·E 3 (ChatGPT), Stable Diffusion (including SDXL, SD3, Pony, Juggernaut, and most forks), Flux.1 and Ideogram, Runway for video, Figma AI for UI scenes, and a generic mode that works in any other tool.
Why does Midjourney need different syntax than DALL·E?+
Midjourney uses weighted comma-separated tags with flags like --ar 16:9 and --stylize 200. DALL·E reads natural prose and ignores tag syntax. Stable Diffusion uses both a positive and negative prompt, usually as comma-separated descriptors. Our generator adapts to each.
What is a negative prompt?+
A negative prompt tells the generator what to exclude — blurry, watermark, extra fingers, signature, text artifacts. Stable Diffusion has a formal NEGATIVE field; Midjourney uses --no; DALL·E responds to 'Avoid: ...' in-line.
Can I use these prompts commercially?+
The prompts themselves are yours. Commercial use of the generated images depends on the tool's license (Midjourney Pro/Standard plan terms, OpenAI DALL·E terms, SD model licenses). Check the specific tool's commercial-use policy.
How do I get better results from these prompts?+
Start with the Balanced variant. If results are bland, try Vivid & painterly. If they're too busy, try Minimalist editorial. Iterate on the subject field with more concrete nouns and verbs, and always use a specific lens/camera rather than just 'photograph'.