AI Image Prompt Writing Tips
## What's Inside This isn't a generic "how to write prompts" guide. It's a structured reference library of 200+ tested prompts organized by three dimensions: artistic style, AI model, and real-world use case. The file is a searchable PDF (optimized for desktop and tablet) containing: - **30+ prompts for Midjourney v6** – covering photorealism, cinematic lighting, architectural visualization, and character design. Includes parameter flags like `--ar 16:9` and `--s 250` with explanations of why they're used. - **25+ prompts for DALL-E 3** – focused on composition control and style adherence. Examples span editorial illustration, product mockups, and surreal landscapes. - **40+ prompts for Stable Diffusion** – split by checkpoint (Realistic Vision, DreamShaper, Deliberate) with negative prompt examples. Includes guidance on CFG scale and sampler selection. - **20+ prompts for Leonardo AI** – optimized for the platform's preset system and Alchemy mode. Covers game assets, concept art, and texture generation. - **A "prompt anatomy" table** – breaks down every component: subject, medium, environment, lighting, color palette, mood, and technical parameters. You'll learn *why* each part matters, not just what to copy. - **Use-case sections** – prompts organized for social media thumbnails, book covers, fantasy character portraits, product photography, architectural renders, and abstract patterns. - **A "what not to do" page** – common mistakes like overstuffing keywords, ignoring aspect ratio, or using contradictory style terms. Each mistake has a before/after example. No fluff. No history of AI art. Just prompts and the logic behind them. ## How to Use This works best as a reference, not a one-time read. Here's a realistic workflow: 1. **Identify your goal.** Are you making a book cover? A social media post? A concept for a client? Flip to the relevant use-case section. 2. **Select your model.** Each prompt is tagged for Midjourney, DALL-E 3, Stable Diffusion, or Leonardo AI. If you use multiple models, you'll find parallel prompts that produce similar outputs on each platform. 3. **Adapt the prompt.** Copy the example, then swap the subject noun and adjust the mood adjective. The structure stays sound. For instance, replace "a weathered lighthouse" with "a neon-lit cyberpunk market stall" and keep the lighting/color parameters. 4. **Tweak parameters.** The guide explains what each parameter does. If you want more stylized output, increase the stylization value. If you need photorealism, lower it. The notes tell you which levers to pull. 5. **Iterate quickly.** Use the "prompt anatomy" table to isolate what changed between a good and bad output. You'll learn to debug your own prompts instead of guessing. This isn't a "type this and get a masterpiece" file. It's a toolkit for getting consistent, usable results faster. ## Who It's For - **Digital artists transitioning to AI** – you already know composition and color theory. This gives you the syntax to express that knowledge in prompt form. - **Content creators who need visual assets** – YouTube thumbnails, blog headers, social media banners. You don't want to spend hours tweaking. You want reliable outputs with minimal fuss. - **Solo entrepreneurs building product visuals** – book covers, mockups, ad creatives. You're not a designer, but you need something that doesn't look generic. - **Writers and storytellers** – you need character portraits, scene references, or map illustrations. The prompts here are structured to generate narrative-driven images, not just pretty pictures. - **Anyone frustrated by inconsistent outputs** – if you've tried writing prompts and gotten random results, this guide explains the systematic reasons why. It's *not* for absolute beginners who have never opened an AI image tool. You should know what Midjourney or Stable Diffusion is and have generated at least a few images. This accelerates you past the trial-and-error phase. ## A Short, Honest Note You'll receive a PDF file – no physical product, no software. Prompting techniques evolve. Models update. What works today on Midjourney v6 might shift with v7. The structural logic in this guide (how to combine subject, style, and technical parameters) is more durable than any individual prompt. But no resource can guarantee specific outputs from generative AI. Regional considerations: If you're using AI tools for commercial work, check the terms of service for each platform in your jurisdiction. Some models have different licensing rules in Europe, Asia, or elsewhere. This guide doesn't cover legal advice – that's your responsibility to verify. No fake testimonials here. I've used these prompts myself across hundreds of generations. Some worked on the first try. Others needed iteration. That's normal. The value is in having a structured starting point so you waste fewer credits and less time. One last thing: if you're outside the US, note that pricing, taxes (like VAT or GST), and payment methods vary by region. The $9 price is in USD; your local equivalent may differ. Gumroad handles the conversion at checkout.
Get it → sunshine4255.gumroad.com