“A pretty landscape, please.”
- No subject details (era, mood, palette)
- No style or DP reference
- No --ar, --stylize, --v parameter flags
- No --no negative prompt
HotPrompt's Midjourney prompt generator emits a single flat, comma-delimited prompt ending in --ar and parameter flags — exactly how Midjourney likes them — with a negative prompt line for clean results.
You'll need an account to optimize. .
Before vs After
Midjourney rewards flat, comma-delimited prompts that end in parameter flags. Skip the prose; lead with the subject; finish with --ar and --no.
“A pretty landscape, please.”
cinematic widescreen landscape of an alpine valley at first light, mist curling between black-pine ridgelines, a glacier-fed river cutting through fields of wildflowers, golden-hour rim light, soft volumetric god rays, shot on Hasselblad H6D, 80mm lens, f/8, deep focus, painterly muted palette of teal mist and warm amber, Roger Deakins grade, ultra detailed, 8K --ar 16:9 --stylize 250 --v 6 --no people, signage, watermark, text, low contrast
How it works
Midjourney prefers a single flat line ending in parameter flags. Four steps to get there without a prompt-engineering PhD.
Subject first, then style, then context — the generator structures the comma-separated order for you.
Lens, lighting, palette, era — described in tags the model recognizes.
Midjourney truncates long inputs. The optimizer fits everything into a clean, parseable line.
--ar, --stylize, --v, --no — sensible defaults you can tweak, baked into every output.
Midjourney wants a single flat line ending in parameter flags — every output respects the format.
Subject → style → context → camera/lens → palette → grade (the order MJ parses best).
Parameter flags appended (--ar, --stylize, --v, --niji when relevant).
--no negative prompt killing common artifacts (warped faces, text, watermarks).
350-char cap so MJ never truncates the prompt mid-clause.
Style references and director-of-photography names that MJ recognizes.
Niji v6 anime mode when the idea is anime — without you having to ask.
Each MJ version has its own personality. The generator picks the right vocabulary so the output composes cleanly.
Specific subject details (breed, age, wardrobe, posture) land better than abstract style words.
MJ ignores 'please make it widescreen' but obeys --ar 16:9. We use parameters everywhere we can.
--no people, signage, text, watermark, low contrast kills 80% of common failures.
When you have an --sref image URL, the generator slots it in correctly so the style transfer applies.
Anywhere a high-quality image saves a stock-photo subscription or a designer's afternoon.
Mood boards, hero shots, reference plates — generated in seconds with --ar matching the brief.
Concept art, tileable textures, character refs — the negative line keeps text and watermarks out.
Book covers, chapter visuals, character portraits — atmospheric grades chosen for the mood you brief.
Landing-page heroes, ad creative, social cards — --ar 16:9 or --ar 9:16 switched in one click.
Questions
Defaults the generator picks: --ar 16:9 for cinematic, --ar 1:1 for square social, --ar 9:16 for vertical. --stylize 250 is a sweet spot — high enough to be artful, low enough to respect the subject. You can always tweak after.
Yes — when your idea mentions 'anime', 'manga' or specific anime references, the generator switches the parameter from --v 6 to --niji 6 and tightens the subject vocabulary for Niji's strengths.
Yes for SD and Flux — the tag-style format is what they want. DALL·E 3 prefers more natural language; for DALL·E targets, the generator switches to descriptive prose with the same subject/style/composition coverage.
Midjourney has historically capped its prompt window around that length. Going longer often gets the trailing instructions truncated. The generator respects the cap so every word counts.
Those are static lists. HotPrompt rewrites your specific idea into a structured tag list — same shape, but actually about what you want. It also adds the --no line, which the random Lexica copy-paste usually skips.
More tools
Ten ready-to-draw words and phrases per roll — with hints, difficulty tags and audience filter.
Five birthday card concepts — AI-image prompt + matching greeting message, tailored to your recipient.
Upload any image — get the prompt that would have generated it. Photographer-grade analysis.