Delivery Apps Are Begging Restaurants to Use AI Photos in 2026 — Here's the Complete Guide to Every Platform's Rules

Restaurants with professional-quality menu photos get 30-35% more orders on delivery platforms. That is not a guess. DoorDash, Uber Eats, and Grubhub have all published data showing the gap between listings with strong photos and listings with phone snapshots or no images at all. Grubhub's own research puts the number even higher: 70% more orders for menus with photos and descriptions versus text-only listings. And yet, most restaurants still upload blurry, dark, incorrectly cropped images that get rejected or buried in the algorithm. The reason is simple: every platform has different specs, different aspect ratios, different content rules, and different review processes. It is a mess. This guide cleans it up. We are covering the six platforms that matter globally: DoorDash, Uber Eats, Grubhub, Deliveroo, Swiggy, and Zomato. Exact specs, content rules, common rejection reasons, and how to produce photos that pass review on all of them without hiring a photographer for each platform.

First, Let's Talk About That Reddit Hoax

In late 2025, a viral Reddit post from a supposed food delivery company "whistleblower" racked up over 87,000 upvotes, spread to X with 36.8 million impressions, and convinced half the internet that delivery platforms were running entirely AI-fabricated restaurant listings. Fake menus, fake storefronts, fake everything. It was a hoax. Journalist Casey Newton used Google's SynthID watermark detection to confirm the "evidence" was AI-generated. The post itself was flagged as 100% AI-written by detection service Pangram Labs. DoorDash CEO Tony Xu publicly denied it, Uber Eats COO Andrew MacDonald called the allegations "completely made up." Why this matters for you: The hoax created confusion between two very different things: Fabricating entire fake restaurants with AI (fraud, against every platform's TOS, and not what anyone should be doing). Using AI to enhance real photos of real food from your real restaurant (legitimate, encouraged by multiple platforms, and increasingly standard practice).

These are not the same thing. Every platform prohibits stock photos and fake listings. No platform prohibits you from improving the lighting, background, or presentation of your actual dishes using AI tools. In fact, Swiggy has partnered with AI photography services to help restaurants do exactly that. The goal is accuracy with appeal: your real food, looking its best. Now let's get into the specs.

Platform-by-Platform Requirements

The Master Comparison Table

| Platform | Menu Item Ratio | Min Resolution | Max File Size | Formats | Review Time | |----------|----------------|----------------|---------------|---------|-------------| | DoorDash | 16:9 (landscape) | 1400 x 800px | 16 MB | JPG, PNG | 1-5 days | | Uber Eats | 5:4 to 6:4 | 1200 x 800px | 10 MB | JPG, PNG | 1-3 days | | Grubhub | 4:3 (landscape) | 1024 x 768px | 10 MB | JPG, PNG | 1-3 days | | Deliveroo | 1:1 (square) | 800 x 800px | 6 MB | JPG, PNG | 1-5 days | | Swiggy | 1:1 (square) | 1024 x 1024px | 5 MB | JPG, PNG | 1-2 days | | Zomato | 1:1 (square) | 800 x 800px | 5 MB | JPG, PNG | 1-3 days |

Notice the problem immediately: three different aspect ratios across six platforms. A landscape shot that looks perfect on DoorDash gets its sides chopped on Deliveroo. A square shot optimized for Swiggy wastes half the frame on Grubhub. This is why a one-size-fits-all approach fails.

DoorDash (US, Canada, Australia, Japan)

Aspect ratio: 16:9 landscape (mandatory). DoorDash crops the center square for thumbnail previews, so your dish must look good both full-width and center-cropped. What gets rejected: Text, graphics, logos, or borders on the image. Partially visible dishes (the entire item must be in frame). Photos that do not match the specific menu item. Stock imagery or photos clearly not of your food. Collages or multiple dishes in one frame. What works: Single dish, landscape orientation, dish filling most of the frame. Clean background, natural-looking lighting. Hero ingredient visible and centered (survives the square crop). Consistent style across menu categories. Pro tip: DoorDash's algorithm favors listings with complete photo coverage. A menu where every item has a photo outperforms one where only "hero items" are photographed. Shoot everything.

Uber Eats (Global — 45+ countries)

Aspect ratio: Between 5:4 and 6:4 for menu items. Cover photos require exactly 2880 x 2304px at 5:4. What gets rejected: Wrong aspect ratio (most common rejection reason). Multiple items in one photo. Text or logos overlaid on the image. Blurry, out-of-focus, or poorly lit shots. Flash-washed images or hard shadows. Stock photos from search engines. What works: One item per photo, centered in frame. Soft, diffused lighting (no harsh flash). Minimal props that do not distract. Accurate color representation. Pro tip: Uber Eats uses AI-based quality scoring internally. Photos with balanced exposure, sharp focus on the food, and accurate colors rank higher in search results within the app. This is not officially documented but has been confirmed by multiple restaurant consultants.

Grubhub (US)

Aspect ratio: 4:3 landscape for menu items, 4:3 for category headers. What gets rejected: Text, coupons, or promotional content on images. Photos of restaurant interiors or exteriors (food only). Images that match URLs in reverse Google image search (they check for stolen photos). Adult, violent, or medical-flagged content. What works: Clean food-only shots on simple backgrounds. Good contrast between dish and background. Original photos (they verify originality). Consistent lighting across the menu. Pro tip: Grubhub's performance data shows that consistent photo quality across an entire menu matters more than having a few standout shots. Invest in making your worst photos better rather than making your best photos perfect.

Deliveroo (UK, Europe, Middle East, Asia-Pacific)

Aspect ratio: 1:1 square for menu items, 16:9 for hero/banner images. What gets rejected: Hands, faces, or people in the frame. Watermarks or branding overlays. Raw ingredients scattered around the dish. Stock photography. Cluttered or distracting backgrounds. What works: Single dish centered in a square frame. Simple, uncluttered background. Ingredients clearly visible. Professional-level lighting and sharpness. Pro tip: Deliveroo operates in markets where visual standards are high (UK, UAE, Singapore, Hong Kong). Restaurants competing in these markets see outsized returns from investing in photo quality because the baseline expectation from consumers is already elevated.

Swiggy (India)

Aspect ratio: 1:1 square. Thumbnail images display at 11:13 (1760 x 2080px) in some placements. What gets rejected: Watermarks, logos, or borders. Stock images or images from other restaurants. Cluttered backgrounds. Low-resolution uploads. What works: Square composition with dish centered. Clean, simple background. Original photos of your actual food. Bright, well-lit presentation. Key detail: Swiggy uses an AI validation system that automatically checks uploaded photos for compliance. If your image fails the automated check, it never reaches a human reviewer. Swiggy has also partnered with Spyne.ai to offer AI-powered photoshoot services directly to restaurant partners, and actively rewards restaurants that meet photo quality thresholds through their accelerator program. This platform is explicitly pro-AI-enhancement.

Zomato (India, UAE, select international)

Aspect ratio: 1:1 square for menu items. Restaurant listing photos must follow a specific sequence: Facade, Ambience, Food. What gets rejected: AI-generated images that fabricate dishes (their policy targets fully synthetic images, not enhancements). Incorrect photo sequence on listings. Low-resolution images that fail mobile optimization. Technical specification non-compliance. What works: Photos of your actual dishes, enhanced for lighting and presentation. Following the Facade > Ambience > Food sequence for listings. High-resolution square crops. Accurate color and portion representation. Key detail: Zomato made headlines in late 2024 by announcing a policy against AI-generated food images. The important nuance: this targets fully fabricated dishes that do not exist in the restaurant's actual menu. Using AI to improve the lighting, remove a messy background, or enhance the color accuracy of a real photo of a real dish you serve is a different category entirely. The distinction is accuracy. If the customer orders the dish and it looks like the photo, you are on the right side of the policy.

The Real Problem: Three Aspect Ratios, Six Platforms, One Kitchen

Here is the math that breaks most restaurants: You have 40 menu items. You need photos for 6 platforms. That is 240 individual image files. In 3 different aspect ratios (16:9, 5:4, 1:1). Each with different minimum resolutions. Each with content rules that vary slightly. Even if you nail the photography, the production and formatting workload is what kills consistency. Restaurants either: Upload one version everywhere and accept that it looks wrong on most platforms. Spend hours in Photoshop cropping and resizing. Skip platforms entirely because the photo requirements are too different. None of these are good options.

How FoodPhoto.ai Handles This

This is the specific problem FoodPhoto.ai was built for. Here is how it maps to the delivery platform challenge:

HD Output That Meets Every Platform's Minimums

FoodPhoto.ai generates images at HD resolution up to 1344px on the long side (1344x768 landscape, 1024x1024 square, or 768x1344 portrait). That exceeds the minimum requirements for every platform listed above: | Platform | Min Required | FoodPhoto.ai Output | Status | |----------|-------------|---------------------|--------| | DoorDash | 1400 x 800 | 1344 x 768 | Meets with minor upscale | | Uber Eats | 1200 x 800 | 1344 x 768 | Exceeds | | Grubhub | 1024 x 768 | 1344 x 768 | Exceeds | | Deliveroo | 800 x 800 | 1024 x 1024 | Exceeds | | Swiggy | 1024 x 1024 | 1024 x 1024 | Meets | | Zomato | 800 x 800 | 1024 x 1024 | Exceeds |

One Photo, Multiple Crops

Upload one base photo of your dish. FoodPhoto.ai enhances it with professional lighting, clean backgrounds, and appetizing presentation. Then export in the aspect ratios you need: 16:9 landscape for DoorDash. 5:4 for Uber Eats. 4:3 for Grubhub. 1:1 square for Deliveroo, Swiggy, Zomato. Same dish, same visual identity, correct specs for every platform.

What AI Enhancement Actually Does (And Does Not Do)

To be specific about what FoodPhoto.ai does, since this is the core question after the Reddit hoax discourse: It does: Improve lighting and exposure on your real dish photos. Clean up or replace messy backgrounds. Enhance color accuracy so the photo matches what the food actually looks like. Generate multiple angles and presentations of dishes you describe. Produce consistent visual style across your entire menu. It does not: Create photos of dishes you do not serve. Fabricate portion sizes larger than reality. Generate fake restaurant listings or storefronts. Produce images intended to deceive customers about what they will receive. The goal is the same as hiring a food photographer: make your real food look its absolute best. The difference is cost, speed, and the ability to do it for every single item on your menu instead of just the top sellers.

Platform-Specific Tips That Actually Move the Needle

For DoorDash: Shoot Wide, Crop Center

Since DoorDash uses 16:9 but crops to square for thumbnails, compose your shots with the dish centered and important details in the middle 60% of the frame. Do not put garnishes or side items at the edges because they will get cut in thumbnail view.

For Uber Eats: Nail the Lighting

Uber Eats has the strictest quality scoring for lighting. Avoid flash entirely. Use window light or a simple diffused LED panel. If your kitchen has no natural light, this is where AI enhancement pays for itself immediately because fixing bad lighting after the fact is exactly what these tools are good at.

For Grubhub: Photograph Every Single Item

Grubhub's data is clear: menus with complete photo coverage dramatically outperform partial coverage. Do not cherry-pick your best dishes. Shoot everything, including the sides, the drinks, and the desserts.

For Deliveroo: Think Square From the Start

If you are on Deliveroo, compose your shots as squares from the beginning. Do not try to crop a landscape shot into a square after the fact because you will lose too much of the dish. Place the plate in the center of a square frame with clean margins on all sides.

For Swiggy: Use Their AI Tools (And Yours)

Swiggy is the most AI-friendly platform. They literally offer AI photography services to partners. Take advantage of their tools for basic shots, then use FoodPhoto.ai for premium enhancement that makes your listing stand out from competitors using the same Swiggy-provided baseline.

For Zomato: Accuracy Is the Policy

Do not get spooked by the "AI ban" headlines. Zomato's policy is about accuracy, not technology. If your enhanced photo accurately represents what the customer will receive, you are compliant. Focus on making real dishes look professional, not on making them look like something they are not.

The Global Opportunity

This is not just a US conversation. Delivery platforms are growing fastest in markets where restaurant photo quality is still low: | Region | Key Platforms | Photo Quality Gap | |--------|--------------|-------------------| | US / Canada | DoorDash, Uber Eats, Grubhub | Medium (improving) | | UK / Europe | Deliveroo, Uber Eats, Just Eat | Medium-High | | India | Swiggy, Zomato | High (massive opportunity) | | Southeast Asia | Grab, Foodpanda, ShopeeFood | Very High | | Latin America | Rappi, iFood, PedidosYa | Very High | | Middle East | Talabat, Deliveroo, Careem | High | In India alone, Swiggy and Zomato together serve over 50 million monthly active users. In Latin America, Rappi and iFood are growing at 30%+ year-over-year. Restaurants in these markets that invest in photo quality now have a disproportionate advantage because the competition has not caught up yet.

The Bottom Line

Every delivery platform wants the same thing: clear, accurate, appetizing photos of real food. They just ask for it in annoyingly different technical specifications. The restaurants winning in 2026 are not the ones with the biggest photography budgets. They are the ones with a system that produces consistent, platform-compliant images across their entire menu and updates them when dishes change. That system used to require a photographer, a photo editor, and someone who understood each platform's specs. Now it requires a phone, decent natural light, and an AI tool that handles the rest. If you have been putting off your delivery app photos because the requirements felt overwhelming, they were. But the tooling has caught up. Start with your top 10 sellers, get them right on every platform, and watch what happens to your order volume.

Your menu deserves better photos

Try 10 photos for $3 or monthly plans from $5 (20 credits). Try 10 photos for $3 → Plans from $3 → View pricing → No commitment. Credits roll over. Cancel anytime.