Midjourney YouTube Thumbnail Prompts: --ar 16:9 Kits

A midjourney youtube thumbnail prompt should lock landscape framing first: end with --ar 16:9 so the composition matches YouTube surfaces before you crop. This guide ships Midjourney v7 kits for face-led, product, split, and mystery layouts, plus text-space and parameter notes for mid-2026. PromptMake writes thumbnail prompt text only. It does not render JPG or PNG files or upload to YouTube Studio. Soft path: https://promptmake.net/youtube-thumbnail-prompt. You leave with kit anatomy, four paste-ready Midjourney kits, finishing rules, common mistakes, and an FAQ. CTR checklist posts and multi-model workflow posts live elsewhere. Stay here for Midjourney --ar 16:9 kits you can paste this week.

Why --ar 16:9 comes first

YouTube thumbnails display as landscape. Generating square or portrait and cropping later cuts faces, hooks, and product edges. Midjourney’s --ar 16:9 sets the canvas so large shapes land where Studio expects them. You still export around 1280x720 in your downloader or editor. The parameter is composition insurance, not a pixel guarantee.

Order the prompt by what must survive phone size: subject and expression first, then scene, light, palette, framing, finish. Append parameters at the end. --ar 16:9 belongs every time. Optional --style raw when Midjourney beautify fights a documentary look. Optional --sref or --cref when series identity matters and you have rights to the references.

Midjourney still invents letters. Treat on-image text as optional and risky. Prefer empty negative space or a blank sign for Canva typography when spelling must be exact. Ideogram owns short quoted hooks when you need generated words; this article stays on Midjourney kits.

Who this fits: creators who already pay for Midjourney and want portable thumbnail language. Who should use a pixel maker instead: people who need a finished file in five minutes with templates and stock faces. Soft prompt planning still helps either path: https://promptmake.net/youtube-thumbnail-prompt.

Kit anatomy for Midjourney thumbnail prompts

Each kit below follows the same sticky-note structure: hook intent, subject clause, contrast pair, text-space instruction, and the Midjourney tail --ar 16:9. Swap nouns; keep hierarchy. One promise per frame. Extra props kill mobile readability.

Hook intent is the viewing decision: reaction, result, object under tension, before-after, or mystery object. Subject clause names who or what sits largest. Contrast pair sets warm skin against cool background, or bright product against dark void. Text-space tells Midjourney where not to pack detail so you can add two to five words later.

Generate three concepts per video that change the hook, not just the adjective. Phone-shrink every candidate. Upload the winner after Canva polish when brand type matters. PromptMake guests get about three generations per day on the youtube-thumbnail-prompt path; free accounts get about five on that path as of mid-2026.

Clause order that survives mobile

Subject and emotion first. Scene second. Light and palette third. Framing fourth. Finish words last. Parameters last of all. If the face is the product, say face scale early: large face left third or right third, shoulders cropped, eye contact.

Text-space without trusting generated letters

Ask for clean negative space upper third or solid sign board with no letters. Add the hook in Canva. If you insist on Midjourney text, keep one short word and expect a redraw. Soft structured drafts: https://promptmake.net/youtube-thumbnail-prompt.

Face-led --ar 16:9 kit

Face-led thumbnails win when expression carries the promise. Lock a large face, clear eyes, and a simple background that does not compete. Avoid forced open-mouth reaction on every upload; match expression to channel tone.

Paste skeleton: [creator description], [emotion], large face filling left third, sharp eye contact, [simple background], high contrast between skin and backdrop, clean empty space on right for title text, cinematic key light, youtube thumbnail composition --ar 16:9

Swap left/right when your series places faces on a fixed side. Keep wardrobe simple. Busy patterns become noise at phone size.

Face kit variants

Curiosity: slight head tilt, raised brow, muted background. Shock: wide eyes, cooler rim light, darker void behind. Confidence: calm half-smile, brighter key, brand-color backdrop.

Face kit Midjourney notes

Use --cref with a face you have rights to when identity must hold across a season. Prompt text alone will drift. Combine PromptMake Image-tab reverse notes with Midjourney character reference when you start from a headshot.

Product and object --ar 16:9 kit

Product channels need the object large and readable. Hands optional. Reflections controlled. Background simplified so the SKU silhouette pops. Phone viewers decide in a glance whether the object is the video’s promise or decoration, so scale beats lifestyle clutter in most Midjourney thumbnail tests.

Paste skeleton: hero [product], large centered or rule-of-thirds placement, [material details], dramatic [light], dark uncluttered background, high contrast edges, empty upper band for text, packshot energy, youtube thumbnail --ar 16:9

Name the tension: cracked screen, glowing power meter, before-box versus open-box. Tension beats beauty-only shots for click decisions. Soft topic-to-prompt expansion when the product hook is still fuzzy: https://promptmake.net/youtube-thumbnail-prompt.

Product kit variants

Unboxing: open carton edge, product rising, warm practical light. Comparison: two objects left-right with a clear size difference. Macro detail: one feature fills half the frame; keep --ar 16:9 so crop still fits Studio.

Product kit Midjourney notes

Logos and fine packaging type will smear. Composite real logos in post when brand accuracy matters. Soft brief expansion for product hooks: https://promptmake.net/youtube-thumbnail-prompt.

Split and mystery --ar 16:9 kits

Split layouts sell transformation. Mystery layouts sell a question the title completes. Both need a hard center divide or a single cryptic object, not eight small puzzles.

Split paste skeleton: split screen youtube thumbnail, left [before state], right [after state], hard vertical divide, matched camera height, high contrast, large shapes, empty lower third for short label space --ar 16:9

Mystery paste skeleton: single mysterious [object] on dark background, strong rim light, extreme scale, no readable text, tension, clean negative space right third, youtube thumbnail --ar 16:9

Keep the title responsible for the verbal hook. The image asks; the title answers, or the reverse. Do not duplicate the full title inside the frame.

Split kit edit rules

Match lens and camera height across both halves or the split looks fake in a distracting way. Change only the story variable: dirty versus clean, empty versus full, sad versus smiling.

Mystery kit edit rules

One object. One light story. If viewers cannot name the question in a second, simplify. Avoid micro clues that die on mobile.

Parameters, finishing, and PromptMake

Minimum Midjourney tail for this job: --ar 16:9. Add --style raw for less default polish on news or documentary channels. Add --stylize only when you measured a need; high stylize can soften faces into beauty filters that hurt trust thumbnails.

Finishing stack: pick the Midjourney still, drop into Canva or Photoshop at 1280x720, add two to five words max, stroke or shadow for contrast, check safe margins, export JPG, upload Studio. PromptMake never skips those steps for you.

Draft kits from a topic line or a reference thumbnail at https://promptmake.net/youtube-thumbnail-prompt. Choose Midjourney as the target when the tool offers model routing. Edit the --ar 16:9 line if missing. Render in Discord or the Midjourney app. You own pixels.

Series habit: save four kits (face, product, split, mystery) with locked face side and palette. Change only the hook noun per episode. Consistency trains returning viewers.

What this page does not cover

Multi-model Ideogram versus FLUX routing belongs to other thumbnail posts on this blog. Generic CTR face-contrast-hook checklists belong to the CTR article. Free pixel-maker lists belong to the free maker route post. Here the deliverable is Midjourney --ar 16:9 kits.

Soft free-tier note

Guests: about three runs per day on the youtube-thumbnail-prompt path. Free accounts: about five per day on that path. Image host fees stay separate. Soft start: https://promptmake.net/youtube-thumbnail-prompt.

Common mistakes with Midjourney thumbnail prompts

Mistake 1: Forgetting --ar 16:9 and cropping faces later.

Mistake 2: Packing five props and a paragraph of invented text.

Mistake 3: Expecting PromptMake to download a JPG.

Mistake 4: Using the same open-mouth face on every niche.

Mistake 5: Changing face side every episode in a branded series.

Mistake 6: Judging full-resolution beauty instead of phone-size clarity.

Mistake 7: Treating Midjourney letters as final brand typography.

Repair path: one subject, one contrast pair, empty text band, --ar 16:9, three hook variants, phone shrink, Canva type, upload.

Series kits and A/B habits

Channels that upload weekly win when thumbnails feel related without cloning. Lock face side, palette family, and typography style in Canva. Change the Midjourney hook noun and expression per episode. Viewers learn where to look; Studio data then tells you which hook family earns clicks.

Run A/B inside YouTube’s tools when you qualify, or rotate concepts across similar videos and compare CTR by traffic source. Do not crown a winner from one viral outlier. Keep a simple log: kit type, hook, Midjourney seed or job link, CTR after seventy-two hours.

PromptMake helps when you need fresh wording for the same kit structure. Feed the topic, keep Midjourney as the target, and preserve your --ar 16:9 and text-space rules in the edit pass. Soft path: https://promptmake.net/youtube-thumbnail-prompt.

Archive losing concepts. A mystery kit that failed on tutorials might win on true-crime. Kits are reusable patterns, not one-shot prompts.

Episode brief template

Four lines: video promise, hook type (face, product, split, mystery), must-show object or face note, words reserved for Canva. Generate Midjourney stills only after those four lines exist.

When to leave Midjourney

Move lettering-heavy hooks to Ideogram or pure Canva type on a plate with empty space. Move exact product packaging to photo composites. Midjourney remains strongest for stylized composition and mood under --ar 16:9.

Soft next steps

Copy the face-led kit. Swap in your channel nouns. Generate three Midjourney stills with --ar 16:9. Shrink them. Finish one in Canva.

When the topic is fuzzy, expand a one-line idea at https://promptmake.net/youtube-thumbnail-prompt, force Midjourney framing, then paste into Midjourney. Remember: prompts, not JPG.

FAQ

What makes a strong midjourney youtube thumbnail prompt?

A strong prompt leads with subject and emotion, sets a clear contrast pair, reserves text space, and ends with --ar 16:9. Keep one promise per frame. Add Midjourney parameters after the visual clauses, not in the middle of the subject line.

Why use --ar 16:9 for YouTube thumbnails?

YouTube displays thumbnails in landscape. --ar 16:9 tells Midjourney to compose for that shape so faces and products land correctly before you export near 1280x720. Cropping from square after the fact often cuts the hook.

Does PromptMake render Midjourney thumbnails?

No. PromptMake writes thumbnail prompt text for Midjourney, FLUX, Ideogram, and similar hosts. You paste into Midjourney and export the image yourself. Soft path: https://promptmake.net/youtube-thumbnail-prompt.

Should I put title text inside the Midjourney prompt?

Prefer empty negative space and add exact type in Canva when spelling matters. Midjourney lettering remains unreliable. If you generate a word, keep it short and proof every character before publish.

How many Midjourney concepts should I make per video?

Three concepts that change the hook beat ten adjective tweaks. Phone-shrink each candidate, pick one, finish type, then upload. Use Studio CTR by traffic source against your own baseline.

Can I keep a consistent face across thumbnails?

Use Midjourney --cref or other character reference features with images you have rights to, and keep face side and crop locked in your kit. Prompt nouns alone will drift across a season.

How do I start on the free tier?

Open https://promptmake.net/youtube-thumbnail-prompt, enter a video topic or upload a reference, and generate Midjourney-oriented prompt text. Guests get about three runs per day; free accounts get about five on that path as of mid-2026. Then render in Midjourney and finish pixels in your editor.