Copy-paste prompt templates organized by category. Each template follows the 7-element architecture and includes tool recommendations, negative prompts, and customization tips.
Updated March 202615+ TemplatesAll Skill Levels
What Are AI Art Prompt Templates?
AI art prompt templates are pre-structured text descriptions that you can copy, paste, and customize in tools like Midjourney, Stable Diffusion, DALL-E, and Flux. Each template uses the proven 7-element prompt architecture (Subject, Environment, Style, Lighting, Mood, Camera, Detail) to ensure high-quality, intentional results. Replace the bracketed placeholders with your own details, add the negative prompt, and generate.
How to Use These Templates
Every template below follows the same format. The prompt text in gold is what you copy into your AI tool. Items in [brackets] are placeholders for you to replace with your own details. The negative prompt in red goes into your tool's negative prompt field (or use Midjourney's --no parameter). Each template also lists which tools produce the best results and tips for customization.
For the best results, read our Prompt Engineering Guide to understand why each element matters. These templates are starting points -- the more you customize them to your specific vision, the better your results will be.
15+
Templates
8
Categories
6
Tools Covered
Portraits
Portrait prompts focus on human (or humanoid) subjects with emphasis on facial detail, expression, and personal character. These templates work best in vertical aspect ratios (2:3 or 4:5) and benefit from shallow depth of field settings.
Template 1: Cinematic Character Portrait
Prompt
A [age] [gender] [character description] with [distinctive features], wearing [clothing/armor details], [expression] expression, [pose], set against [background], cinematic portrait photography, 85mm lens f/1.4 shallow depth of field, [lighting type] lighting with [color temperature] tones, [mood] atmosphere, 8K, ultra-detailed skin texture, shot on Hasselblad X2D
Negative: cartoon, anime, deformed face, extra fingers, blurry, watermark, text, bad anatomy, crossed eyes, low quality, plastic skin
Best tools: Midjourney (--ar 2:3 --s 750), Stable Diffusion (SDXL), Flux. Tip: Specify exact eye color and hair texture for consistency. Add "catchlight in eyes" for lifelike portraits.
Template 2: Fantasy Character Portrait
Prompt
Portrait of a [race/species] [class] with [magical features like glowing eyes, runes on skin, ethereal hair], wearing [ornate armor/robes with specific materials], [expression] gaze, [magical effects around them], dark fantasy illustration style inspired by [artist reference], dramatic rim lighting with [magical color] glow, painterly brushwork, mysterious and powerful mood, bust shot, detailed costume design, 4K
Negative: photorealistic, modern clothing, simple background, low quality, blurry, bad anatomy, extra limbs, watermark
Best tools: Midjourney (--ar 2:3 --s 800), Stable Diffusion (with fantasy LoRAs), Leonardo. Tip: Reference specific fantasy IPs for style anchoring: "Dungeons and Dragons concept art" or "Magic: The Gathering card art quality."
Landscapes
Landscape templates emphasize environment, depth, and atmosphere. Use wide aspect ratios (16:9 or 21:9) for maximum impact. Lighting and mood carry these compositions more than any other category.
Template 3: Epic Natural Landscape
Prompt
[Landscape type] at [time of day], [foreground elements] in the foreground, [main feature] dominating the midground, [distant elements] on the horizon, [weather/atmosphere], [art style] style, [lighting description] casting [shadow behavior], [mood] mood with [color palette] palette, panoramic wide shot, 24mm lens deep depth of field, 8K, ultra-detailed vegetation and terrain textures, National Geographic quality
Best tools: Midjourney (--ar 16:9 --s 800), Stable Diffusion (SDXL), DALL-E. Tip: Time of day dramatically changes the feel. Golden hour and blue hour produce the most cinematic results. Add "mist in valleys" for depth.
Template 4: Fantasy/Sci-Fi Environment
Prompt
A [fantastical location] with [impossible architecture or natural formations], [unique atmospheric element like floating islands, crystal formations, bioluminescent flora], [small figures or creatures for scale], digital matte painting style inspired by [artist: Craig Mullins, Dylan Cole, etc.], [dramatic lighting description] with volumetric god rays, sense of [epic scale/wonder/mystery], extreme wide establishing shot, concept art quality, 8K, trending on ArtStation
Best tools: Midjourney (--ar 21:9 --s 750), Stable Diffusion (SDXL), Leonardo. Tip: Include a small human figure for scale -- it transforms "nice landscape" into "jaw-dropping vista." Specify three distinct depth layers.
Concept Art
Concept art templates are designed for game developers, film pre-production, and worldbuilding. They prioritize clarity of design, readability at different scales, and professional presentation.
Template 5: Character Design Sheet
Prompt
Character design sheet of a [character description], multiple views showing front, side, and back, [costume/armor details with materials and colors], [weapon or accessories], clean white background, concept art style, professional character turnaround, detailed annotations, consistent proportions across all views, studio lighting, flat colors with subtle shading, game-ready design, 4K
Negative: background scene, single view only, photorealistic, blurry, low quality, inconsistent design, watermark, text overlay
Best tools: Midjourney (--ar 16:9), Stable Diffusion (with concept art models), Leonardo. Tip: Add "T-pose" or "A-pose" if you need the character in a neutral stance for 3D modeling reference. Specify "no perspective distortion" for clean reference sheets.
Template 6: Creature/Monster Design
Prompt
Creature concept art of a [creature type] with [unique biological features], [size reference compared to human], [surface texture: scales, chitin, fur, bioluminescent patches], [movement type: quadrupedal, flying, aquatic], in its natural [habitat type], scientific illustration meets fantasy art style, dramatic [lighting] revealing form and texture, sense of [danger/beauty/alien intelligence], three-quarter view, concept art quality, highly detailed anatomy, 4K
Negative: cute, cartoon, chibi, simple, flat, low detail, blurry, watermark, human features
Best tools: Midjourney (--ar 4:3 --s 600), Stable Diffusion (SDXL), Leonardo. Tip: Reference real-world biology for believable creatures. "Deep sea anglerfish anatomy mixed with praying mantis limbs" gives the AI strong visual anchors for hybrid designs.
Product Mockups
Product mockup templates generate clean, commercial-grade product visualizations. These prioritize accurate materials, professional lighting, and aspirational context. Use square or 4:5 aspect ratios.
Template 7: Luxury Product Shot
Prompt
[Product type] with [material finish: brushed metal, matte ceramic, glossy glass], [color] colorway, placed on [surface: marble slab, concrete pedestal, fabric drape], [contextual props: plants, water droplets, fabric swatches], high-end product photography, studio lighting with [key light angle] key light and soft fill, [mood: sophisticated, playful, technical] mood, [background: gradient, solid, textured], medium close-up, 100mm macro lens, f/8 deep focus, commercial photography quality, 8K, clean and minimal composition
Best tools: Midjourney (--ar 1:1 --s 400), DALL-E, Adobe Firefly. Tip: Keep backgrounds simple. One hero product, one surface, minimal props. Add "catalog photography" or "e-commerce product photo" for commercial clarity.
Template 8: Tech Product Lifestyle Shot
Prompt
[Tech device] in a [lifestyle setting: modern workspace, coffee shop, outdoor adventure], [person interacting naturally with the product], [environment details that suggest target demographic], lifestyle product photography, natural window light mixed with [warm/cool] ambient, aspirational and authentic mood, 35mm lens f/2.8, shallow depth of field with product in sharp focus, Apple-campaign level quality, 8K, color graded with [warm/cool/neutral] tones
Negative: studio background, floating product, stock photo feel, fake smile, low quality, blurry, watermark, text on screen
Best tools: Midjourney (--ar 4:5 --s 500), DALL-E, Flux. Tip: Specify the brand aesthetic you want to match without naming the brand. "Minimalist Scandinavian tech aesthetic" or "premium Japanese industrial design" gives strong style direction.
Book Covers
Book cover templates prioritize visual storytelling, mood, and leaving space for title text. Always use vertical aspect ratios (2:3) and consider where text will be placed in the composition.
Template 9: Fantasy Novel Cover
Prompt
[Hero/heroine description] standing [heroic pose] before [epic backdrop: towering fortress, stormy sky, magical portal], [signature weapon/artifact glowing with energy], epic fantasy book cover illustration, dramatic [lighting] with [color] magical illumination, sense of destiny and adventure, large open sky area at top for title text, rich saturated colors, painterly illustration style inspired by [Michael Whelan, Larry Elmore, etc.], vertical composition with strong focal point at center-bottom, ultra-detailed, 4K
Best tools: Midjourney (--ar 2:3 --s 800), Stable Diffusion (SDXL), Leonardo. Tip: Always leave the top 25-30% of the image as relatively simple sky or background for the book title. Add "negative space at top" to encourage this.
Template 10: Thriller/Mystery Cover
Prompt
[Ominous scene: lone figure on rain-soaked street, shadowy corridor, isolated house on cliff], [key visual motif: shattered mirror, red door, trail of blood, flickering streetlight], noir thriller atmosphere, high contrast with deep blacks and [single accent color: red, amber, blue], [fog/rain/snow] obscuring details, cinematic photography style, dramatic chiaroscuro lighting from [single source], tense and foreboding mood, vertical composition with open space for title text at top, 35mm film grain, desaturated palette, 4K
Best tools: Midjourney (--ar 2:3 --s 600), Stable Diffusion (SDXL), Flux. Tip: Thriller covers work best with a single dominant color accent against a dark, desaturated background. Red is classic, but blue creates cold tension and amber creates paranoia.
Social Media Thumbnails
Thumbnail templates are designed for maximum visual impact at small sizes. They use bold compositions, high contrast, saturated colors, and clear focal points that read instantly. Use 16:9 for YouTube or 1:1 for Instagram.
Template 11: YouTube Thumbnail Background
Prompt
[Eye-catching scene or object] with [vibrant, saturated colors], [dramatic effect: explosion, energy burst, magical glow, speed lines], bold and dynamic composition with strong [color] background, [visual element that creates depth], pop art meets cinematic style, extremely high contrast, vibrant [primary color] and [complementary color] color scheme, [dramatic lighting], energetic and attention-grabbing mood, wide shot with clear center focus area for text overlay, 4K, sharp and clean
Best tools: Midjourney (--ar 16:9 --s 400), DALL-E, Adobe Firefly. Tip: Thumbnails need to work at 320x180 pixels. Use bold, simple shapes and no more than 2-3 dominant colors. Leave the center and right third open for face cam or text overlays.
Template 12: Instagram Post Visual
Prompt
[Aesthetically pleasing subject] in a [Instagram-worthy setting], [cohesive color palette matching brand colors: specify 2-3 colors], styled and curated look, [flat lay / lifestyle / portrait] composition, soft natural lighting with [warm/cool] tones, aspirational and shareable mood, clean and polished aesthetic, square composition with balanced elements, professional social media photography, 4K, color graded in [warm earth tones / cool pastels / vibrant editorial]
Best tools: DALL-E, Midjourney (--ar 1:1 --s 500), Adobe Firefly. Tip: Consistency matters for social feeds. Define your 3-color palette once and use it across all posts. Add "cohesive brand aesthetic" to reinforce visual unity.
NFT Art
NFT art templates focus on originality, visual distinctiveness, and collection-ready aesthetics. These should feel unique and ownable. Square format (1:1) is standard for most NFT platforms.
Template 13: Generative Collection Avatar
Prompt
[Character type] avatar with [unique trait: cybernetic eye, crystal horns, plasma hair], [specific outfit/accessory layer: leather jacket with neon piping, ornate collar, holographic visor], [distinct background: solid gradient, abstract pattern, minimal environment], digital art style with [flat color blocks / detailed rendering / pixel art], clean outlines, [signature color palette: neon and dark / pastel and white / earth and gold], collectible character design, centered bust composition, consistent art style suitable for 10K collection, 4K, sharp lines
Best tools: Midjourney (--ar 1:1 --s 600), Stable Diffusion (with consistent style LoRA), Leonardo. Tip: For collections, lock the seed and only change trait variables. Use img2img in Stable Diffusion with a base template to maintain structural consistency across pieces.
Template 14: 1/1 Fine Art NFT
Prompt
[Abstract or surreal concept] merging [unexpected element A] with [unexpected element B], [artistic technique: double exposure, glitch art, collage, fluid dynamics], [rich texture: oil paint, marble, liquid metal, crystalline], transcendent and thought-provoking composition, gallery-quality fine art, [art movement influence: Surrealism, Futurism, Abstract Expressionism], [lighting that enhances the concept], emotionally resonant [specific mood], [color palette that supports the emotional theme], square composition, museum-quality detail, 8K, signed artist piece quality
Best tools: Midjourney (--ar 1:1 --s 1000), Stable Diffusion (SDXL with artistic models), Flux. Tip: 1/1 NFTs should feel like gallery pieces. Push for unusual concept combinations that create visual tension. "A cathedral made of frozen ocean waves, illuminated by bioluminescent deep-sea creatures" is more compelling than "abstract colorful art."
Wallpapers
Wallpaper templates create images designed to live behind desktop icons and mobile interfaces. They need visual breathing room, balanced compositions that work with UI elements overlaid on top, and enough resolution for large displays.
Template 15: Desktop Wallpaper
Prompt
[Atmospheric scene: vast landscape, cosmic vista, serene nature, abstract gradient], [primary focal point positioned off-center to avoid icon overlap], generous negative space across [left side / right side / both sides] for desktop icons, [calming/inspiring] mood with [sophisticated color palette], [art style: photorealistic, digital painting, minimalist, abstract], soft [lighting type] creating gentle gradients, no harsh contrasts or busy details in icon zones, ultrawide composition, smooth tonal transitions, 5K resolution (5120x2880), desktop wallpaper optimized
Negative: text, watermark, busy center, cluttered, high contrast patterns, people in center, distracting details on left side, low quality, blurry, JPEG artifacts
Best tools: Midjourney (--ar 16:9 --s 700), Stable Diffusion (SDXL at high resolution), Flux. Tip: Desktop wallpapers need to work with icons on the left side (Mac) or anywhere (Windows). Keep the left third relatively simple. Sunsets, nebulae, misty mountains, and abstract gradients are perennial favorites because they have natural visual breathing room.
Template 16: Mobile Wallpaper
Prompt
[Vertical scene: tall architecture, waterfall, night sky, forest path], focal point positioned in the center-lower third (avoiding notification bar and dock), [elegant color palette with 2-3 dominant colors], [art style], soft gradients at top for clock/notification readability, [gentle lighting] without harsh highlights, [serene/energetic/mystical] atmosphere, vertical composition optimized for mobile screens, smooth tonal transitions, no small busy details, 4K mobile resolution (2160x3840), AMOLED-friendly [rich blacks / deep colors]
Negative: text, watermark, horizontal composition, busy top area, harsh white areas, low quality, blurry, cluttered, landscape orientation
Best tools: Midjourney (--ar 9:16 --s 700), Stable Diffusion (SDXL), DALL-E. Tip: For AMOLED screens, include "deep true black areas" in your prompt to take advantage of OLED's perfect blacks. This also saves battery. Keep the top quarter dark or simple for clock readability.
Video Tutorial
Watch this walkthrough to see how to use and customize these prompt templates in real-time with multiple AI tools.
Frequently Asked Questions
Copy the prompt text (shown in gold) directly into your AI art tool. Replace everything in [brackets] with your own specific details. Paste the negative prompt (shown in red) into your tool's negative prompt field -- in Midjourney, convert it to a --no parameter. Adjust tool-specific settings like aspect ratio, stylize value, or CFG scale as recommended in each template's tips section.
Each template lists which tools produce the best results. Most work across Midjourney, Stable Diffusion, DALL-E, and Flux with minor adjustments. Midjourney users should append parameters like --ar and --s. Stable Diffusion users should utilize the separate negative prompt field and tune CFG scale (7-12 recommended). DALL-E users can simplify the prompt to more natural, conversational language while keeping the core descriptors.
Yes, these templates are designed to be customized. Replace the bracketed placeholders with your own subjects, settings, and preferences. Change the lighting, mood, and camera settings to match your vision. You can also mix elements from different templates. The underlying 7-element structure ensures quality regardless of your specific modifications.
Start with the Landscape templates or the Cinematic Character Portrait. Landscapes are forgiving because the AI excels at natural environments, and small imperfections are less noticeable. The portrait template gives you a strong structure to follow. As you gain confidence, move to Concept Art and NFT Art categories, which require more precise prompt control. Read our Prompt Engineering Guide for foundational knowledge.
For consistency, use seed control: lock the seed number and change only one variable at a time. Always include negative prompts to prevent unwanted drift. Keep aspect ratio, quality settings, and sampler/stylize values the same. In Midjourney, consistent --s values help. In Stable Diffusion, keep the same sampler (DPM++ 2M Karras is reliable) and CFG scale. Save your best prompts in a text file and iterate on them rather than starting fresh.
Common aspect ratios: Portraits use 2:3 or 4:5 (vertical). Landscapes use 16:9 or 21:9 (cinematic wide). Social media squares use 1:1. Instagram stories and TikTok use 9:16. YouTube thumbnails use 16:9. Book covers use 2:3. Desktop wallpapers use 16:9 or 16:10. Mobile wallpapers use 9:16. NFT art typically uses 1:1. Product shots use 1:1 or 4:5.