10 AI Photo Prompts for Photorealistic Results

The difference between an AI-generated image that looks like a photo and one that actually looks like a specific kind of photograph lies almost entirely in the prompt. Photography is a technical discipline with precise vocabulary — camera systems, lens choices, lighting setups, exposure relationships, film stocks, compositional traditions — and AI image models have been trained on enormous quantities of real photography. When your AI photo prompts speak that vocabulary precisely, the results cross the line from ‘looks like a photo’ to ‘looks like a specific photographer shot this on a specific camera in specific light’.

Below are 10 prompts across 10 photography genres — street photography, documentary, macro, landscape, architectural, sports action, food photography, underwater, astrophotography, and medical/scientific. Each includes the full prompt, a breakdown of the specific photographic choices and why they matter, and guidance on adapting the approach for your own subjects and creative goals.

A photograph is not just a picture — it is the product of specific decisions: the camera and lens, the exposure triangle (aperture, shutter speed, ISO), the light source and its quality, the compositional approach, and the photographer’s editorial eye. When you write ‘a photo of a busy street’, the AI makes all of those decisions by averaging what it has seen. When you specify the camera, lens, aperture, light, and photographic tradition, the AI makes the same decisions a photographer would — and the result looks like a photograph, not a rendering.

Claude is particularly useful for building photo prompts because it understands photographic language deeply. Describe the shot you want — the scene, the mood, the genre — and ask Claude to translate that into a technically specific photo prompt with the camera settings, light quality, and photographic reference that will produce the result you are reaching for.

Use case: editorial photography, documentary content, urban lifestyle imagery, social media street photography aesthetic.

Candid street photograph, Tokyo Shibuya crossing at rush hour, shot from street level looking up slightly, wide angle 28mm lens, deep depth of field showing entire crossing filled with pedestrians, overcast diffused daylight, wet pavement reflecting neon signs and pedestrian signals. Black and white conversion, high contrast, deep shadows, bright highlights on wet surfaces. Grain texture of Ilford HP5 pushed to 3200 ISO. Decisive moment — all pedestrians mid-stride. Documentary photojournalism quality, Henri Cartier-Bresson compositional tradition.

What makes this work: ‘shot from street level looking up slightly’ is the specific camera position that gives street photography its immersive quality. ‘Decisive moment’ is Cartier-Bresson’s foundational concept and naming it tells the AI to capture a peak-action frame rather than a static pose. ‘Ilford HP5 pushed to 3200 ISO’ specifies a real film stock at a specific push-process level, producing the characteristic gritty grain that differentiates authentic film-style street photography from a clean digital conversion. The wet pavement instruction doubles the visual interest through reflection.

Adapt it by: changing the city and crossing type, the lens focal length (50mm for a more compressed, intimate feel), the time of day (night adds neon colour; dawn adds emptiness and solitude), and the film stock reference (Kodak Tri-X for a slightly different grain character).

Use case: editorial portrait photography, documentary journalism, profile imagery, NGO and development content.

Documentary portrait of an elderly fisherman mending nets on a weathered wooden dock at dawn, face deeply lined and weathered, looking down at his work, not posing. Available light from the low dawn sun creating warm sidelighting on his face. Shot on 35mm lens from 1.5 metres, f/2.8 creating shallow depth of field that softens the dock and ocean behind while keeping his hands and face sharp. Warm film colour grade, slight desaturation in shadows, skin tones prioritised. Magnum Photos documentary quality, Steve McCurry colour tradition.

What makes this work: ‘not posing’ is one of the most important instructions for documentary portrait AI photography — without it, AI defaults to the direct gaze, formal pose of a studio portrait. The specific distance (‘1.5 metres’) and aperture (f/2.8) produce a specific depth of field relationship that is the visual signature of mid-range documentary portrait work. ‘Skin tones prioritised’ tells the model to protect warm accurate skin rendering in the colour grade. Naming Steve McCurry gives the model a specific colour signature — warm, rich, saturated but not oversaturated.

Adapt it by: changing the subject, their occupation and environment, the time of day and its light quality, the focal length and distance (longer focal length compresses the relationship between subject and environment), and the documentary photographer reference.

Use case: nature content, scientific illustration, product photography, fine art nature prints, editorial close-up work.

Macro photograph, extreme close-up of a single honeybee on a lavender flower head, shot at 1:1 magnification, 100mm macro lens. Ultra-shallow depth of field — only the bee’s compound eye and immediate face in sharp focus, wings and body already softening, lavender florets completely dissolved into a soft purple and green bokeh. Diffused morning sidelight with catchlight visible in the compound eye. Every individual facet of the compound eye visible. Shot on full-frame sensor, f/5.6. National Geographic natural history photography quality.

What makes this work: specifying ‘1:1 magnification’ establishes the exact scale of macro photography — the subject is reproduced at life size on the sensor. The depth of field description (‘compound eye sharp, wings already softening, lavender completely dissolved’) gives the model a precise three-plane focus map that produces authentic macro depth of field. ‘Catchlight visible in the compound eye’ is the specific detail that makes macro insect photography feel alive — the tiny reflection of the light source in the eye. ‘Every individual facet of the compound eye visible’ establishes the level of detail expected.

Adapt it by: changing the insect and host plant, the magnification ratio (2:1 or greater for even more extreme close-up), the focus plane (thorax instead of eye for a different compositional emphasis), and the background colour through the bokeh of the plant behind.

Use case: fine art landscape prints, nature and travel photography, desktop wallpapers, editorial landscape content.

Long exposure landscape photograph, rocky coastal scene at blue hour, 4-minute exposure smoothing the ocean into a silky mist around the rocks, clouds streaking across the deep blue sky in motion blur arcs. Rocks rendered with absolute sharpness due to tripod stability contrasting with the ethereal motion in water and sky. Colour palette of deep navy, steel grey, pale cyan, with warm amber light from distant horizon. Shot on 16-35mm lens at 18mm, f/11, ND filter stack. Perfectly level horizon. Fine art landscape photography quality, Michael Kenna long exposure tradition.

What makes this work: naming the exact exposure time (‘4-minute exposure’) produces the specific degree of motion smoothing that differentiates a 4-minute from a 30-second exposure — the former reduces moving water to near-glass, the latter to soft texture. ‘Rocks rendered with absolute sharpness contrasting with the ethereal motion’ describes the specific visual tension that makes long exposure landscape photography compelling. ‘Blue hour’ is a precise time of day — approximately 20-30 minutes after sunset — with its characteristic deep blue ambient light. ‘Perfectly level horizon’ is a technical quality marker. Michael Kenna is the definitive reference for long exposure minimalist landscape photography.

Adapt it by: changing the landscape type (mountain lake, forest waterfall, urban cityscape at night), the exposure time, the time of day (golden hour vs blue hour vs full night), and the focal length and compositional approach.

Use case: architecture and design publications, real estate content, interior design, brand photography for built environment clients.

Architectural interior photograph, brutalist concrete library interior, dramatic atrium space with coffered ceiling casting geometric shadow patterns across the floor, students studying at long wooden tables below. Shot with tilt-shift lens correcting vertical convergence for perfectly parallel verticals. Ambient natural light from high clerestory windows, no flash. Colour: cool grey concrete, warm wood, soft natural light. Shot at f/8 for maximum depth of field, medium-format aesthetic of extreme tonal detail. Ezra Stoller architectural photography tradition. Technically immaculate, spatially powerful.

What makes this work: ‘tilt-shift lens correcting vertical convergence for perfectly parallel verticals’ is the specific technical device of professional architectural photography — without it, wide-angle shots of tall interiors produce converging verticals that look amateur. This single instruction signals professional architectural photography to the model. ‘Coffered ceiling casting geometric shadow patterns’ is a specific compositional detail that turns an architectural element into a graphic composition. Ezra Stoller is the foundational reference for mid-century American architectural photography — his work defined how architecture is photographed for print.

Adapt it by: changing the architectural style and period (modernist, classical, industrial), the interior function (gallery, station, chapel), the light source quality and direction, and the time of day and its effect on the space.

Use case: sports editorial, event photography, sports brand content, news and magazine sports sections.

Sports action photograph, professional basketball player at peak of jump shot, ball leaving fingertips at the top of the arc, arena crowd blurred in background. 1/2000th second shutter speed freezing all motion completely, water droplets of sweat suspended mid-air. Shot on 200mm telephoto at f/2.8, shallow depth of field isolating player from background. Stadium floodlight top lighting creating rim light on player. Getty Images sports photography quality, dramatic, peak moment captured.

What makes this work: ‘water droplets of sweat suspended mid-air’ is the specific detail that proves the shutter speed claim — at 1/2000th second, sweat droplets are genuinely frozen in space, and including this as a visible element tests and confirms the technical specification. The combination of 200mm telephoto at f/2.8 produces the specific compression and shallow depth of field that sports photography is characterised by — a compressed, isolated subject against a soft, impressionistic crowd. ‘Stadium floodlight top lighting creating rim light’ describes the actual lighting condition of indoor arena sports.

Adapt it by: changing the sport and its characteristic peak moment, the venue and its lighting quality, the shutter speed (motion blur can be intentional in some sports photography), and the focal length (wider angle for environmental context, longer for compression and isolation).

Use case: restaurant and food brand content, cookbook photography, editorial food journalism, menu imagery.

Professional food photograph, freshly plated pasta carbonara in a wide shallow bowl on a worn linen tablecloth, steam wisping from the surface, golden egg yolk pooled in the centre with black pepper and crispy pancetta. Shot overhead (flat lay) from exactly 90 degrees. Soft north-facing window light from the left, no shadows. Shot on 100mm macro lens at f/8 for full depth of field. Colour: warm ivory linen, golden egg yolk, the rich off-white sauce. Deliberate styling with imperfect edges suggesting just-plated freshness. Editorial food photography quality, dark and moody Nordics food aesthetic.

What makes this work: ‘from exactly 90 degrees’ specifies the overhead flat lay precisely — even a few degrees off produces unwanted perspective distortion in food flat lays. ‘Soft north-facing window light from the left, no shadows’ describes the specific diffused natural light setup that professional food photographers most commonly use. ‘Deliberate styling with imperfect edges suggesting just-plated freshness’ is the specific food styling instruction that differentiates editorial food photography from the overly perfect, clinical look. ‘Dark and moody Nordics food aesthetic’ is a recognisable contemporary food photography genre with a specific colour and shadow language.

Adapt it by: changing the dish and its visual components, the camera angle (45-degree hero shot vs overhead flat lay), the lighting direction and quality, the surface and prop styling, and the colour palette and mood direction.

Use case: nature and wildlife photography, conservation content, diving and ocean brand imagery, scientific documentation.

Underwater photograph, large manta ray gliding directly overhead from below, full wingspan visible, diver swimming alongside for scale, photographed looking up toward the light-filtered surface. The surface above creates a shimmering backlit silver-blue ceiling, manta’s wing edges lit by the light from above. Shot on wide 16mm rectilinear lens to show full wingspan, f/8 for depth of field. Deep ocean blue, warm filtered light from above, white belly of the manta creating strong contrast against the blue. National Geographic underwater photography quality, Brian Skerry visual tradition.

What makes this work: ‘photographed looking up toward the light-filtered surface’ is a specific camera angle — shooting upward in underwater photography produces the characteristic silvery backlit surface effect that is visually distinctive. ‘The surface above creates a shimmering backlit silver-blue ceiling’ describes that effect explicitly. Including a ‘diver swimming alongside for scale’ is the compositional device that communicates the genuine enormity of the manta ray — without a human reference point, size is abstract. Brian Skerry is one of the most recognisable underwater photographers for National Geographic.

Adapt it by: changing the marine subject, the camera angle (horizontal shot vs looking up vs looking down), the depth and its effect on colour (deeper water shifts toward monochromatic blue), the lighting source (strobe flash for vivid colour vs available light for atmosphere), and the scale reference.

Use case: space and science content, nature fine art prints, editorial science photography, desktop wallpapers.

Astrophotography landscape, the Milky Way core rising above an ancient stone circle in a moorland landscape, 25-second exposure at ISO 6400, stars rendered as points not trails, the galaxy’s central bulge visible with subtle nebulosity and dark lanes. Foreground lit by a single dim torch held by a lone figure standing at the stones, warm orange light painting the granite. Colours: deep blue-purple sky, the yellow-orange core of the Milky Way, cold blue star field, warm orange torch light. Shot on 14mm ultra-wide, f/1.8. Babak Tafreshi Milky Way landscape tradition.

What makes this work: ‘stars rendered as points not trails’ is a critical technical instruction — at 25 seconds, stars should be at the limit of point rendering before trailing begins, and specifying this ensures the AI produces the correct star sharpness rather than long trails. ‘A single dim torch held by a lone figure’ provides both scale and the warm-cold colour contrast that makes Milky Way landscape photography so visually compelling. The specific colour description for each zone (blue-purple sky, yellow-orange core, cold blue star field, warm orange torch) gives the model a complete colour map. Babak Tafreshi is the most recognisable photographer of night sky landscapes.

Adapt it by: changing the foreground element (mountain peak, desert dune, coastal cliff), the light painting source and colour, the exposure parameters (longer exposures for star trails, shorter for sharper stars), and the season and Milky Way orientation.

Use case: scientific communication, medical and health editorial, educational content, research publication imagery.

Scientific scanning electron microscope photograph (SEM), surface of a common salt crystal (sodium chloride), extreme magnification showing the perfect cubic crystalline structure at microscopic scale, individual crystal faces and edges with remarkable geometry, surface imperfections and growth steps visible. False colour applied — warm gold and amber tones applied to crystal faces, deep charcoal shadow in recesses. Scale bar visible in corner. Shot at 500x magnification. The kind of image published in Nature or Scientific American. Technically perfect, scientifically accurate, visually beautiful.

What makes this work: ‘scanning electron microscope photograph (SEM)’ names a specific imaging technology with a characteristic visual signature — the deep focus, the three-dimensional surface rendering, the particular quality of electron microscope imagery. ‘False colour applied — warm gold and amber tones’ describes the specific post-processing technique used to make SEM images more visually accessible and beautiful, since true SEM images are inherently greyscale. ‘Scale bar visible in corner’ is the specific authenticity marker that every real SEM image carries. Specifying magnification (500x) gives the model a sense of scale.

Adapt it by: changing the subject and its characteristic microscale structure (pollen grain, butterfly wing scale, spider silk, circuit board), the magnification and the detail level it reveals, the false colour palette, and the imaging technology (light microscopy vs SEM vs TEM produces very different visual signatures).

Every prompt in this collection specifies camera position and angle (what the camera is doing, not just what it is pointing at), lens focal length and aperture (which controls compression, depth of field, and spatial relationships), light source quality and direction (the single most important variable in photography), shutter speed where relevant (which controls motion representation), film stock or colour grade, and a named photographer or publication tradition. These are the variables that make a photograph look like it was shot by a professional who made deliberate choices — not assembled by a model defaulting to the average of everything it has seen.

Use Claude to build photo prompts before generating. Describe the shot in plain language — what you want to show, the mood, the genre — and ask Claude to translate that into a technically specified photo prompt with camera settings, light quality, and photographic tradition. Save the prompts that produce the strongest results in Chat Smith as reusable templates.

The most common mistake is describing the subject without describing the photograph. ‘A photo of a mountain’ tells the model the subject. ‘Long exposure landscape, mountain at dawn, graduated ND filter, 16mm, f/11, blue hour light, golden alpenglow on the summit’ tells the model the photograph. The second most common mistake is forgetting camera position and angle — where the camera is relative to the subject (shooting up, down, level, close, far) is one of the most powerful variables in photography and one of the most reliably neglected in AI photo prompts.

A third mistake is using generic quality markers instead of technical specifications. ‘High quality’, ‘professional photography’, ‘stunning’ — these add almost nothing. ‘Shot on medium format, f/8, tilt-shift corrected verticals’ — these add everything. The model does not respond to quality claims; it responds to technical specifications that imply quality through the choices being described.

A great AI photo prompt is a great shot brief. It specifies the camera system and lens, the exposure relationship, the light quality and direction, the camera position, and the photographic tradition it belongs to. These 10 AI photo prompts demonstrate that across 10 completely different photographic genres. Take any of them, adapt the technical variables to your specific subject and vision, iterate with targeted refinements, and the output will cross the line from ‘looks like a photo’ to ‘looks like this specific photographer made this specific shot’.

Building great AI photo prompts is a technical skill that compounds with practice. Chat Smith lets you save your best photo prompts as one-click templates organised by genre or project, use Claude to translate rough shot ideas into technically specified prompts, compare the same prompt across multiple AI image models to find which handles a specific photographic genre most convincingly, and build a personal photo prompt library that captures everything you have learned about what produces authentic photographic results.

You can also ask Claude to generate a suite of prompt variations — the same scene shot at different times of day, with different lenses, in different photographic traditions — to test a full range of options before committing to a generation session. The technical briefing stage is where the most significant quality gains in AI photography happen.

1. Which AI tools produce the most realistic photography results?

Midjourney v6 and later versions produce exceptionally photorealistic results with detailed prompts, particularly for natural light and outdoor photography. DALL-E 3 handles compositional precision and specific content very accurately. Stable Diffusion with photorealistic checkpoints (such as Realistic Vision or PhotoReal) offers the most controllable photographic output. Adobe Firefly is strong for commercial photography simulation. The prompts in this collection are written to work across tools — the photographic vocabulary is understood universally. Test the same prompt across multiple tools to find which handles your specific genre most convincingly.

2. How do I make AI photos look less digitally generated?

Three things matter most. First, specify film grain or a real film stock — the organic imperfection of grain is the most powerful signal of authentic photography. Second, avoid the direct, centred, slightly too-perfect composition that AI defaults to by specifying camera angle, off-centre framing, and compositional rules (rule of thirds, leading lines). Third, add candid or environmental authenticity — subjects not posing, environmental details that would only be present in real situations, slight imperfections in the scene that signal it was not constructed for the camera.

3. Can I specify real camera brands and models?

Yes, and it is often useful. ‘Shot on Leica M10’ produces a different aesthetic signal than ‘shot on Canon 1DX’ or ‘shot on Hasselblad medium format’. AI models have absorbed the associations between camera systems and photographic genres from the vast amount of photography discussion and metadata they have been trained on. Leica signals street and documentary photography; Hasselblad signals fashion and fine art; Canon 1DX signals sports and wildlife. The camera body reference is a shorthand for a whole cluster of aesthetic expectations.

4. How do I maintain consistent photographic style across multiple images?

Build a photographic style template — a fixed block of text that specifies your consistent aesthetic choices: camera system, focal length range, aperture preference, colour grade, film stock, and photographer reference. Append this to every subject-specific prompt. The subject changes; the style template remains constant. Save the template in Chat Smith so it is always available as a one-click addition. Consistent style across a body of AI photography comes from consistent prompt language, not from any model memory. The technical specification is the style guide.