Social platforms each have their own canvas, their own attention economy, and their own taste. A great Instagram post is a terrible YouTube thumbnail. This tutorial gives you the platform-specific prompt patterns that produce graphics designed for actual feeds — not orphan images that look good in isolation but flop when published.
Social-media graphics differ from artistic AI images in two ways. First, the canvas is fixed — and getting the aspect ratio wrong means cropping disasters. Second, almost every social graphic ends up with overlaid text added in Figma or Canva. That means your AI image should be composed with empty negative space where the text will sit. This tutorial covers both — the aspect-ratio playbook and the text-friendly composition trick.
Think of every social graphic as a layered cake. The bottom layer is the AI-generated background. The middle layer is a typography overlay added in Canva or Figma. The top layer is platform UI (likes, captions, the algorithm thumbnail crop). Your prompt has to think about all three layers from the start — especially by leaving a clean, low-contrast area for the text to live.
--ar 1:1 (1080×1080)--ar 4:5 (1080×1350)--ar 9:16 (1080×1920)--ar 16:9 (1280×720)--ar 1.91:1 (1200×627), Midjourney accepts --ar 1200:627--ar 4:1 (1584×396)--ar 2:3 (1000×1500)Most social graphics need a clean low-contrast area for an overlaid headline. Prompt for this explicitly: "composition leaves the upper-left third as a soft uncluttered area suitable for typography overlay". The model will obligingly place the subject in the right two-thirds and leave you a clean rectangle for your headline. This single trick saves more re-rolls than any other.
YouTube thumbnails live or die by three things: a clear focal point readable at tiny sizes, high colour contrast against the platform's white-and-grey UI, and one strong emotional cue (a face, a bold object, a moment of action). Prompt for: "high-contrast YouTube thumbnail composition, bold colour blocking, single clear focal point in the right two-thirds, expressive face or hero object, vibrant saturation, readable at small sizes".
Generic image — composed without platform in mind
a beautiful photo of a person at a laptop, for my InstagramNo aspect ratio — output will be the default square. No negative-space planning — when you try to add a headline in Canva, it lands on top of the subject's face. No emotional cue, no contrast strategy. The image will look generic and underperform on the feed.
Platform-aware prompt with text-overlay planning
YouTube thumbnail composition for a video titled
"My First Year Freelancing — Honest Numbers".
Hero element: a 30-something Indian woman, expressive
laughing-but-tired face, sitting at a sunlit home desk,
placed in the right two-thirds of the frame.
Composition: upper-left third left intentionally clean
and uncluttered with soft warm-pastel wall behind her,
for a large yellow text overlay added later in Canva.
Lighting: warm window light from the left, soft and flat
enough not to compete with overlay text.
Style: photorealistic, slightly oversaturated for thumbnail
visibility, 35mm lens, shallow depth of field, contemporary
YouTube creator aesthetic.
--ar 16:9 --v 6 --style raw --s 200 --no text, watermark,
deformed hands, captionsThe output is thumbnail-ready: a clear emotional face in the right two-thirds, a clean soft pastel area in the upper left waiting for the headline, and a colour palette that pops against YouTube's neutral grey UI. Drop the title in Canva and you have a thumbnail that earns clicks.
Tip: Build a personal "platform pack" template — one prompt structure per format (square post, vertical reel, YouTube thumbnail, LinkedIn banner). Plug-and-play for any campaign, and your weekly content production time drops to a fraction.
Pick a topic you would post about. Generate a square Instagram post, a vertical Reel cover, and a 16:9 YouTube thumbnail — all for the same content idea. Notice how dramatically composition has to change across formats.
Generate a YouTube thumbnail with the negative-space trick (clean upper-left, hero on the right). Add a real headline in Canva and compare its readability to a thumbnail without the negative space.
Try DALL·E 3 for an in-image headline (e.g. "SALE 30% OFF"). Compare to Midjourney's attempt at the same. This solidifies which tool to reach for when text is part of the brief.
--ar before anything else.Sign in to join the discussion and post comments.
Sign inPrompt Engineering for Business & Productivity
Use AI to work smarter — automate tasks, make better decisions, and communicate professionally. 12 practical business prompt tutorials for professionals.
Prompt Engineering for Developers
Use AI as your coding co-pilot. 18 tutorials on writing prompts to generate clean code, debug faster, write tests, build APIs, and ship better software.
Prompt Engineering for Data Science & Analytics
Supercharge your data workflows with AI. 15 practical tutorials on using prompt engineering for data cleaning, EDA, machine learning, SQL, visualisation, and more.
Prompt Engineering for Education & Learning
Use AI as your personal tutor. Learn how to study faster, create lesson plans, generate practice questions, master languages, and prepare for competitive exams with smart prompts.
Advanced Prompt Engineering Techniques
Master the powerful techniques AI experts use every day. Chain-of-thought, RAG, agents, function calling, prompt evaluation, and much more — 20 deep-dive tutorials.
Prompt Engineering for Content & Copywriting
Write blogs, ads, emails, and social media content ten times faster with AI. 13 practical tutorials on prompt engineering for content creators and copywriters.