Photorealistic DALL·E 3 Prompts, Updated for GPT Image (2026)
- Editorial status
- Verified
- Last verified
- 2026-10-15
- Sources
- 1 listed
On this page
Key takeaways
- The photo prompt formula
- Portraits
- Street and documentary
- Products
Advantages and limitations
Quick tradeoff checkAdvantages
- Sentence-style photo prompts carry over from DALL·E 3 to GPT Image and other models
- Camera, light and imperfection details do most of the work
Limitations
- DALL·E 3 itself is no longer available in OpenAI's API
- OpenAI's image model lineup keeps changing; check which model you are using
Status, October 2026: OpenAI removed the dall-e-2 and dall-e-3 models from its API on May 12, 2026, after announcing it on November 14, 2025. OpenAI's deprecations page lists gpt-image-2, gpt-image-1 and gpt-image-1-mini as replacements, and some GPT Image models have their own removal dates (for example gpt-image-1 on October 23, 2026). Check that page before you build on a specific model.
The prompts themselves still work. Full sentences that describe a photograph carry over from DALL·E 3 to GPT Image and to other modern generators. Every prompt below is written that way, with a copy button.
The photo prompt formula
A [shot type] photo of [subject doing something] in [place]. [Light]. Shot on [lens / camera feel], [depth of field]. [Color and film look]. [Small real-world details].
| Part | Examples |
|---|---|
| Shot type | close-up portrait, candid street photo, overhead product shot, wide landscape photo |
| Light | soft window light, overcast daylight, golden hour backlight, single bare bulb |
| Lens | 35mm, 50mm, 85mm portrait lens, 100mm macro, 24mm wide-angle |
| Depth of field | shallow depth of field, background softly blurred, everything in focus |
| Color / film | natural colors, muted tones, slight film grain, Kodak Portra-like warmth |
| Real details | visible skin texture, flyaway hairs, scuffs on the table, condensation on the glass |
The last row matters most. Real photos are full of small imperfections; images without them look like renders.
Portraits
A close-up portrait photo of a 70-year-old fisherman with a weathered face and a grey beard, standing on a harbor at dawn. Soft overcast light. Shot on an 85mm lens, background softly blurred. Natural colors, visible skin texture and pores, a few water droplets on his jacket.
A candid photo of a young woman laughing at a cafe table by a large window. Soft daylight from the side. 50mm lens, shallow depth of field. Muted warm tones, slight film grain, a half-finished cup of coffee in the foreground.
A professional headshot of a man in his forties wearing a navy blazer and white shirt, plain light grey background. Soft studio lighting from the front left. 85mm lens. Natural skin tones, no retouching look.
An environmental portrait of a potter at her wheel in a cluttered studio. Light from a single high window. 35mm lens, clay dust in the air, clay on her hands and apron.
Street and documentary
A candid street photo of a busy night market in the rain. Neon signs reflect on wet pavement, people with umbrellas pass by in motion blur. 35mm lens, high contrast, slight grain.
A documentary-style photo of a train conductor leaning out of an old train window at a rural station. Late afternoon light. 35mm lens, faded colors, dust on the window frame.
Products
A studio product photo of a matte black wireless headphone on a light grey surface. Soft box lighting from above, gentle shadow underneath. 100mm lens, everything in focus, clean minimal background.
A lifestyle product photo of a ceramic coffee mug on a wooden desk next to an open notebook. Morning window light, steam rising from the mug. 50mm lens, shallow depth of field.
An overhead flat-lay photo of a leather wallet, keys, sunglasses and a watch on a linen cloth. Soft even daylight. Everything in focus, neutral colors.
Food
A close-up photo of a bowl of ramen with a soft-boiled egg, sliced pork and green onions. Steam rising, side light from a window. 50mm lens, shallow depth of field, dark wooden table, small drips of broth on the rim.
An overhead photo of a rustic breakfast table: sourdough toast, butter, berries and coffee. Soft morning light, crumbs on the table, natural colors.
Places
A wide landscape photo of a mountain lake at sunrise. Mist over the water, first light on the peaks. 24mm lens, everything in focus, natural colors, a small wooden dock in the foreground.
An interior photo of a small Scandinavian living room. Soft afternoon light through sheer curtains, a wool blanket on a grey sofa, plants by the window. 24mm lens at eye level, natural colors.
An architectural photo of a concrete brutalist building against a clear blue sky. Hard midday sunlight, strong shadows. 24mm lens, vertical lines kept straight.
Animals
A wildlife photo of a red fox sitting in fresh snow. Low winter sun from behind, glowing rim light on its fur. 400mm telephoto lens, background completely smooth.
A close-up photo of a sleeping tabby cat curled up in a patch of sunlight on a wooden floor. 50mm lens, shallow depth of field, individual whiskers visible.
Tips for realism
- Describe a photo, not quality. "Shot on a 50mm lens in soft window light" helps more than a pile of words like "ultra realistic, 8k, masterpiece".
- Add one or two imperfections. Dust, scuffs, flyaway hairs, condensation, uneven light.
- Name the light source. "Light from a single window on the left" is more realistic than "beautiful lighting".
- Write exclusions as sentences. "No text or logos anywhere in the image" works in sentence-based models.
- Keep one lens and one light. Mixed camera and lighting cues give flat, average images. The lens prompt cheat sheet and the weather and time-of-day prompts have more options.
Using these prompts elsewhere
The same prompts work in Midjourney, FLUX and Stable Diffusion. In tools with a negative prompt field, add a short list such as cartoon, 3d render, plastic skin, oversaturated to push the result toward photography; see the negative prompt lists. For SDXL settings, see the SDXL prompt guide.
Build your prompt
The prompt generator helps you combine a subject with camera, lighting and style options. It runs in your browser and does not generate images.
Primary sources
Try the Visual Prompt Generator
Build Midjourney, DALL-E, and Stable Diffusion prompts without memorizing parameters.
Go →See more AI prompt guides
Explore more AI art prompt tutorials and walkthroughs.
Go →Ready to create your own prompts?
Try our visual prompt generator - no memorization needed!
Try Prompt Generator