Midjourney Prompts
The artist-favorite image and video model — beauty and mood first, with a parameter language all its own.
About Midjourney
Midjourney V7 is the latest generation of the model that built the largest artist community in AI imagery. Accessible through both its web app and Discord, it generates images with a signature sense of composition and atmosphere, and can now bring those stills to life as short animated clips.
Its enduring strength is aesthetics. Where other models chase literal accuracy, Midjourney chases beauty — lighting that flatters, palettes that cohere, compositions that feel authored rather than assembled. For mood boards, concept art, album covers, and anything where the vibe is the deliverable, it remains the reference point many artists judge everything else against.
Midjourney also has a prompt culture of its own: compact, evocative phrasing plus parameters like --ar and --sref that steer format and style. Learning that dialect is half the craft, which is why the gallery below collects tested Midjourney prompts with their real outputs — copy one that already speaks the language, then make it yours.
How to write Midjourney prompts
- 1Set the frame before anything else with --ar: wide ratios for cinematic scenes and landscapes, tall ones for posters and character portraits. Composition follows the canvas.
- 2Use --sref with an image URL to lock a visual style across a series — every generation inherits the reference's look, which is how artists keep a whole project coherent.
- 3Tune --stylize to decide who is in charge: lower values keep the model literal to your words, higher values hand more control to Midjourney's own aesthetic instincts.
- 4Lead with mood and medium, not object lists: "quiet dusk, muted teal palette, 35mm film still, soft grain". Midjourney rewards evocative art-direction vocabulary.
- 5Keep prompts short and let the model interpret. Midjourney thrives on suggestion — a tight poetic line often beats a paragraph of clauses.
- 6For video, nail the still first: generate and refine the perfect frame, then animate it, describing one simple motion — drifting fog, a slow turn of the head — so the mood survives the movement.
Frequently asked questions
What is Midjourney V7 best at?
Aesthetics and atmosphere. It excels at images that feel art-directed — coherent palettes, expressive lighting, strong composition — which is why it is the go-to for concept art, mood boards, cover art, and stylized illustration, with animation extending those stills into motion.
What do --ar and --sref do in Midjourney prompts?
--ar sets the aspect ratio of the canvas, shaping composition from the start. --sref points at a reference image whose style every generation will inherit, keeping a series visually consistent. They are the two parameters worth learning first, and the tested prompts on this page show both in context.
Can Midjourney V7 make videos?
Yes — Midjourney can animate its images into short clips. The reliable workflow is stills-first: perfect the frame in image mode, then animate it with a brief description of the motion you want, keeping it to one movement per clip.
Where can I use these Midjourney prompts?
Midjourney runs on its own web app and through its Discord server. Copy any prompt from this page, paste it into either entry point, and adjust the subject or parameters to fit your project.
Related models
MiniMax H3
MiniMax H3 is a powerful multimodal video generation model that creates up to 15-second 2K videos with native stereo sound from text, image, video, and audio inputs.
Seedance 2.0
ByteDance's #1-ranked video model — multi-shot direction with native lip-sync and spatial audio.
GPT Image 2
OpenAI's top-ranked image model — it plans the layout, reasons, and self-checks before it draws, nailing complex multi-part instructions.
Seedance 2.5
Seedance 2.5 is a next-gen AI video model that generates continuous, synchronized 30-second 4K clips from a single prompt.