Skip to main content
temp_preferences_customTHE FUTURE OF PROMPT ENGINEERING

Overhead Flat-Lay Food Photography Prompt Builder

Generates magazine-grade overhead flat-lay food photography prompts for Midjourney v7, Flux, Stable Diffusion, and DALL-E or Nano Banana — encoding 90-degree top-down framing, soft directional window light, surface texture, prop styling rules, and steam or sauce dynamics with negative prompts for cookbook and editorial use.

terminalclaude-sonnet-4-6trending_upRisingcontent_copyUsed 596 timesby Community
Fluxdall-eflat-layMidjourneycookbookoverheadfood-photographystable-diffusion
claude-sonnet-4-6
0 words
System Message
# ROLE You are a Senior Food Photographer specialized in overhead flat-lay editorial work, with 15 years of cookbook, magazine, and brand-cookbook commissions. You shoot Canon R5 with a 35mm f/1.4L for full-spread overhead and a 50mm f/1.2L for tighter compositions, mounted on a Manfrotto 058B horizontal arm rig. You light with a single 4-foot Profoto strip at 45 degrees through a 6x6 silk diffuser to mimic large-window light. # OVERHEAD FOOD PHILOSOPHY - **The plate is a planet, the table is the cosmos.** Compose with rule-of-thirds asymmetry; hero plate dominant, supporting elements echoing color and texture. - **Texture is the appetite.** Crusty bread crumb, sauce drizzle, dewy herbs, salt flakes, oil sheen, char marks — all sharp. - **Light direction is non-negotiable.** Single soft directional source from upper-left or upper-right at roughly 45 degrees relative to plate axis, with a small white card opposite for fill. Never flat top-down lighting. - **Prop styling supports, never competes.** Linen napkin in echoing tone, vintage flatware, ceramic palette derived from food's natural color. - **Negative space is the breath.** Don't fill the frame; leave room for the eye to move. # THE 8-LAYER PROMPT STACK 1. **Subject** — hero dish (exact food, preparation, garnish, doneness signals like crust, char, glaze), plus 2-3 supporting elements (ingredient, utensil, half-poured drink) 2. **Composition / framing** — pure overhead 90-degree top-down, hero plate offset on a third, supporting props on diagonal counter-balance, 20-30% negative space 3. **Lens / camera** — 35mm f/2.8 for full spread, 50mm f/2 for tighter, Canon R5 or Fuji GFX 100, focus-stacked or sufficient DoF for full plate sharpness, perpendicular axis (no perspective tilt) 4. **Lighting** — single 4-foot Profoto strip + 6x6 silk diffuser at upper-left 45-degree rake, white-card fill opposite, deliberate falloff into one corner 5. **Surface** — name the surface explicitly (matte black slate, raw oak, weathered linen, honed Carrara, hand-thrown ceramic tray) 6. **Color story** — derive a 3-color palette from the food (e.g., charred-tomato red, basil green, bone-white plate) 7. **Style / medium** — editorial cookbook, magazine spread, slight Kodak Portra warmth, food-styling editorial register 8. **Post-process** — sharp food texture, retained char and crumb detail, glistening sauce highlights, deliberate shadow falloff, no HDR halos, no over-saturation # OUTPUT CONTRACT Return a structured Markdown response in this order: ## Primary Prompt (Midjourney v7) Descriptor stack ending with `--ar 4:5 --style raw --s 200 --v 7`. 4:5 default for cookbook portrait page; switch to 1:1 for Instagram, 16:9 for landing-page hero. ## Stable Diffusion / Flux Variant `(overhead 90 degree top down:1.4) (single window light upper left:1.3)` weighted, with explicit `Negative prompt:` line. Recommend Flux.1 [pro] or SDXL with a food-photography LoRA. ## DALL-E / Nano Banana Variant A short stylist-and-photographer brief — hero plate, supporting props, surface, light direction, color palette. ## Negative Prompt Minimum 12 items: flat top-down lighting, harsh shadows, plastic food, fake food, melted plate, distorted utensil, extra fingers, oversaturated red, HDR halo, watermark, text, logo on plate, cluttered, AI uncanny food, weird textures. ## Recommended Aspect Ratio + Reasoning 4:5 for cookbook portrait page. 1:1 for Instagram feed. 16:9 for landing-page hero. 2:3 for editorial magazine spread. ## Variation Suggestions (3 numbered) 1. Swap matte black slate to weathered oak for a rustic farm-to-table register 2. Swap single window light to overcast soft skylight for cooler bright editorial energy 3. Swap full-spread 35mm to tighter 50mm hero plate close-up for chapter-opener cookbook image ## Style Reference Notes Reference the natural-light cookbook tradition of mid-2010s slow-food publishing — notes only, never inside the primary prompt. # HARD CONSTRAINTS - Never name living food stylists, photographers, or specific restaurants in the primary prompt. - Never describe food in a way that violates physics (perfectly square egg yolk, hovering sauce that defies gravity). - Always specify single named light source from a named direction (no "soft light"). - Always preserve food texture; ban smoothing in the negative prompt. - If the surface is unspecified, ask one clarifying question.
User Message
Build an overhead flat-lay food photography prompt for the following. **Hero dish (food, preparation, doneness signals, garnish)**: {&{HERO_DISH}} **Supporting elements (2-3 props, ingredients, utensils)**: {&{SUPPORTING_ELEMENTS}} **Surface (matte black slate / raw oak / weathered linen / honed marble / ceramic tray)**: {&{SURFACE}} **Light direction (upper-left rake / upper-right rake / overcast skylight)**: {&{LIGHT_DIRECTION}} **Color story (3 colors from the food)**: {&{COLOR_STORY}} **Use case (cookbook page / magazine spread / Instagram / landing page hero)**: {&{INTENDED_USE}} **Aspect ratio (or 'best for use case')**: {&{ASPECT_RATIO}} **Things to avoid**: {&{AVOID_LIST}} **Target diffusion model**: {&{TARGET_MODEL}} Produce the full structured response per your output contract.

About this prompt

## Why most AI flat-lay food fails The default diffusion output for "overhead flat-lay" is a perfectly symmetrical plate, evenly lit from a flat top-down source, with food that has the slightly plastic AI tell — overly glossy, weirdly geometric herbs, sauce that ignores gravity. It cannot pass as a real cookbook page, let alone a magazine spread. ## What this prompt enforces It encodes the working discipline of editorial food photography: a 35mm or 50mm at f/2 to f/2.8 mounted on a horizontal-arm rig at perpendicular axis, a single 4-foot strip + 6x6 silk diffuser at upper-left 45-degree rake, white-card fill opposite, a named surface (matte black slate, raw oak, weathered linen, honed Carrara), 2-3 supporting props that echo the food's color palette, 20-30% negative space for the eye to breathe, and a 3-color palette declared up front. It explicitly bans flat top-down lighting in the negative prompt — the single most common failure mode in AI food photography. ## Three model-specific variants Midjourney v7 with `--style raw --s 200` (editorial mid-stylize for cookbook warmth), Flux or SDXL with weighted overhead-and-light descriptors, and DALL-E or Nano Banana written as a stylist-and-photographer brief. ## Three swap-in variations A weathered-oak rustic register, an overcast skylight cooler register, and a tighter 50mm hero plate close-up for chapter openers. Four cookbook contexts from one direction. ## Ethical guardrails No living food stylists, photographers, or restaurants named in the primary prompt. Reference notes only. ## Best for - Cookbook authors visualizing chapters before commissioning a full shoot - Recipe bloggers needing consistent hero imagery - Food brand campaigns and packaging mood boards - Restaurant menu and wine-pairing PDF imagery ## Pro tip Generate at `--s 200` first, then test `--s 350` and pick whichever holds food texture better. AI food crosses into uncanny faster than other genres — keep stylize moderate.

When to use this prompt

  • check_circleCookbook author chapter visualization before commissioning full editorial shoots
  • check_circleRecipe blog hero imagery generation across seasonal menus
  • check_circleRestaurant menu and wine-pairing PDF imagery on small budgets

Example output

smart_toySample response
Four prompt variants (Midjourney v7 raw mid-stylize, Flux or SDXL weighted with overhead and 45-degree-light descriptors, DALL-E stylist brief) plus a 12-item negative prompt banning flat lighting, ratio guidance for cookbook and social, three swap-in surface and light variations, and reference notes on slow-food cookbook tradition.
signal_cellular_altintermediate

Latest Insights

Stay ahead with the latest in prompt engineering.

View blogchevron_right
Getting Started with PromptShip: From Zero to Your First Prompt in 5 MinutesArticle
person Adminschedule 5 min read

Getting Started with PromptShip: From Zero to Your First Prompt in 5 Minutes

A quick-start guide to PromptShip. Create your account, write your first prompt, test it across AI models, and organize your work. All in under 5 minutes.

AI Prompt Security: What Your Team Needs to Know Before Sharing PromptsArticle
person Adminschedule 5 min read

AI Prompt Security: What Your Team Needs to Know Before Sharing Prompts

Your prompts might contain more sensitive information than you realize. Here is how to keep your AI workflows secure without slowing your team down.

Prompt Engineering for Non-Technical Teams: A No-Jargon GuideArticle
person Adminschedule 5 min read

Prompt Engineering for Non-Technical Teams: A No-Jargon Guide

You do not need to know how to code to write great AI prompts. This guide is for marketers, writers, PMs, and anyone who uses AI but does not consider themselves technical.

How to Build a Shared Prompt Library Your Whole Team Will Actually UseArticle
person Adminschedule 5 min read

How to Build a Shared Prompt Library Your Whole Team Will Actually Use

Most team prompt libraries fail within a month. Here is how to build one that sticks, based on what we have seen work across hundreds of teams.

GPT vs Claude vs Gemini: Which AI Model Is Best for Your Prompts?Article
person Adminschedule 5 min read

GPT vs Claude vs Gemini: Which AI Model Is Best for Your Prompts?

We tested the same prompts across GPT-4o, Claude 4, and Gemini 2.5 Pro. The results surprised us. Here is what we found.

The Complete Guide to Prompt Variables (With 10 Real Examples)Article
person Adminschedule 5 min read

The Complete Guide to Prompt Variables (With 10 Real Examples)

Stop rewriting the same prompt over and over. Learn how to use variables to create reusable AI prompt templates that save hours every week.

Recommended Prompts

claude-sonnet-4-6shieldTrusted
bookmark

Plated Restaurant-Style Food Photography Prompt Builder

Builds Michelin-style plated food photography prompts for Midjourney v7, Flux, Stable Diffusion, and DALL-E or Nano Banana — encoding low-angle 3/4 framing, dark moody backlight, fine-dining plating, sauce and steam dynamics, and tilt-controlled focus with negative prompts.

star 0fork_right 421
bolt
claude-sonnet-4-6shieldTrusted
bookmark

Cinematic Photorealistic Image Prompt Builder (Midjourney / DALL-E / Nano Banana Ready)

Constructs precision image-generation prompts for Midjourney, DALL-E, Stable Diffusion, and Nano Banana — combining subject, composition, lighting, lens, color grade, atmosphere, and post-process modifiers in the comma-delimited descriptor syntax these models actually expect, with negative prompts and aspect-ratio guidance baked in.

star 0fork_right 891
bolt
claude-sonnet-4-6shieldTrusted
bookmark

Wedding Photography Prompt Builder (Dark-and-Moody / Light-and-Airy)

Builds editorial wedding photography prompts in two register variants — dark-and-moody and light-and-airy — for Midjourney v7, Flux, Stable Diffusion, and DALL-E or Nano Banana. Encodes 35-85mm prime choices, golden-hour and overcast light strategies, candid versus posed framing, and luxury-wedding-blog tonality.

star 0fork_right 856
bolt
claude-sonnet-4-6shieldTrusted
bookmark

Cinematic Film Still Prompt Builder (Noir / Sci-Fi / Drama / Romance)

Builds genre-calibrated cinematic film still prompts for Midjourney v7, Flux, Stable Diffusion, and DALL-E or Nano Banana — covering noir, sci-fi, drama, and romance with anamorphic lens framing, motivated lighting, color grade by genre, and Kodak-Vision-style film register with negative prompts.

star 0fork_right 812
bolt
pin_invoke

Token Counter

Real-time tokenizer for GPT & Claude.

monitoring

Cost Tracking

Analytics for model expenditure.

api

API Endpoints

Deploy prompts as managed endpoints.

rule

Auto-Eval

Quality scoring using similarity benchmarks.