Camera words tell an AI image tool how close the photo is, how blurred the background is and from where the picture is taken. This guide explains lens sizes, aperture, camera angles and crop terms in plain language, gives you copy-paste camera recipes, and includes five free prompts for ChatGPT, Gemini, Midjourney and Stable Diffusion.
Table of Contents
Key Takeaways
- A good camera part names four things: lens size, aperture, angle and crop.
- 85mm at f/1.8 is the classic flattering portrait look. 35mm is wider and more environmental.
- These words describe a look. The AI does not use a real lens, so results are close, not exact.
- Use one lens and one angle per prompt. Mixed instructions give muddy results.
Why Camera Words Matter
Without camera words, the AI chooses its own framing, usually a centered, evenly sharp, slightly wide shot. That is the most common look of a generic AI image. A few specific words change distance, depth and perspective, and they are part four of our 5-part prompt formula.
If you are new to prompts, read the complete beginner guide first. Camera words work best together with clear lighting, so also read our lighting words cheat sheet.
Note: An AI image tool does not have a real lens. Words like "85mm" work because the tool has learned what photos taken with that kind of lens look like. Use them as strong hints, not exact settings.
Lens Terms Explained
The lens size, in millimetres, sets how much of the scene is in the picture and how the perspective looks.
| Lens words | Look | Best for | Watch out |
|---|---|---|---|
| 14-24mm ultra-wide | Huge field of view, strong perspective | Landscapes, architecture, dramatic scenes | Stretches and distorts faces near the edges |
| 28-35mm wide | Wide, includes the surroundings | Street, travel, environmental portraits | Slight face distortion if the camera is very close |
| 50mm standard | Natural, close to how the eye sees | Everyday portraits, candid shots | Less background blur than longer lenses |
| 85mm portrait | Flattering face, soft background blur | Portraits and headshots | Needs distance, so the crop is tighter |
| 105-135mm short telephoto | Compressed background, strong blur | Beauty and close portraits | Background looks closer to the subject |
| 200mm+ telephoto | Strong compression, very flat perspective | Sports, wildlife, distant subjects | Rarely right for casual portraits |
Aperture and Background Blur
Aperture is written as an f-number. A small f-number means a wide opening, which gives a shallow depth of field: the subject is sharp and the background is blurred.
| Words | Effect |
|---|---|
| f/1.4 or f/1.8 | Very shallow focus, strongly blurred background, glowing out-of-focus lights |
| f/2.8 | Shallow focus with a little more of the subject sharp |
| f/5.6 | Moderate focus, a softly blurred but recognisable background |
| f/8 to f/11 | Most of the scene sharp, good for landscapes and groups |
Tip: Add "sharp focus on the eyes" with any wide aperture. Otherwise the AI may blur the face instead of the background.
Camera Angle Terms
The angle sets the viewer's relationship to the subject. Choose one.
| Angle words | Where the camera is | Feeling |
|---|---|---|
| eye-level | Level with the subject's eyes | Honest, natural, connected |
| low angle | Below the subject, looking up | Powerful, heroic, tall |
| high angle | Above the subject, looking down | Small, gentle, vulnerable |
| bird's-eye view / top-down | Directly overhead | Pattern, layout, flat-lay food or objects |
| worm's-eye view | On the ground, looking up | Dramatic, strong sense of height |
| Dutch angle | Camera tilted sideways | Tension and unease, for cinematic scenes |
| over-the-shoulder | Behind one person, looking past them | Conversation and storytelling |
Crop and Shot Size Terms
The crop decides how much of the person is in the frame. Name it directly, because tools often default to a medium shot.
| Crop words | What is in frame |
|---|---|
| extreme close-up | Eyes, lips or one small detail |
| close-up | Face only, from forehead to chin |
| head-and-shoulders | Face and shoulders, the classic headshot |
| waist-up / medium shot | Person from the waist up |
| cowboy shot | Person from mid-thigh up |
| full-length shot | The whole body from head to feet |
| wide shot | Person small in a large scene |
Camera Recipes for Common Goals
Copy a recipe into the camera part of your prompt and change the details to fit your subject.
| Goal | Camera words to paste |
|---|---|
| Flattering portrait | 85mm lens, f/1.8, eye-level, head-and-shoulders crop, sharp focus on the eyes |
| Natural candid photo | 50mm lens, f/2, eye-level, waist-up shot, slightly off-centre composition |
| Heroic or powerful look | 35mm lens, f/2.8, low angle, full-length shot |
| Travel or street scene | 28mm lens, f/4, eye-level, wide shot with the surroundings sharp |
| Dreamy beauty close-up | 105mm lens, f/2, eye-level, close-up, softly blurred background |
| Food or flat-lay | 50mm lens, f/5.6, top-down view, even soft light |
5 Copy-Paste Prompts
Each prompt uses a different camera setup. Replace the example subject with your own, then run the same subject with a different lens and angle to see the difference.
1. Flattering Portrait (85mm)
Prompt 01
Best for: ChatGPT · Gemini · Midjourney
A woman in her thirties wearing a cream knit sweater, gentle smile, looking slightly off camera [subject], realistic portrait in a quiet green park [style], soft golden-hour sunlight from camera right [lighting], 85mm lens, f/1.8, eye-level, head-and-shoulders crop, sharp focus on the eyes, creamy blurred background [camera], warm natural tones, natural skin texture [mood]. Free from plastic skin, extra fingers and warped text. Add a small watermark text reading 'www.PromptsForPhotos.Com' in the bottom-right corner, small clean sans-serif font, pure white, 50% opacity, unobtrusive, not overlapping the subject.
2. Heroic Low-Angle Shot (35mm)
Prompt 02
Best for: ChatGPT · Gemini · Midjourney
A man in his forties in a charcoal coat standing with his arms crossed, calm confident expression, looking at the camera [subject], realistic photograph on a city rooftop at sunrise [style], warm low sunlight from camera left [lighting], 35mm lens, f/2.8, low angle, full-length shot, buildings visible behind him [camera], strong and calm, natural colours, natural skin texture [mood]. Free from distorted hands, warped text and extra limbs. Add a small watermark text reading 'www.PromptsForPhotos.Com' in the bottom-right corner, small clean sans-serif font, pure white, 50% opacity, unobtrusive, not overlapping the subject.
3. Wide Street Scene (28mm)
Prompt 03
Best for: ChatGPT · Gemini
A young woman in a yellow raincoat walking along a wet cobblestone street, laughing, holding a paper cup of tea [subject], candid street photograph in a small old town [style], soft overcast daylight [lighting], 28mm lens, f/4, eye-level, wide shot showing shopfronts and the street, everything reasonably sharp [camera], cheerful and calm, muted natural colours, natural skin texture [mood]. Free from plastic skin, extra fingers and warped text. Add a small watermark text reading 'www.PromptsForPhotos.Com' in the bottom-right corner, small clean sans-serif font, pure white, 50% opacity, unobtrusive, not overlapping the subject.
4. Beauty Close-Up (105mm)
Prompt 04
Best for: Midjourney · Gemini · ChatGPT
A close-up portrait of a woman in her twenties with freckles and light eyes, looking directly at the camera, relaxed neutral expression [subject], beauty portrait against a soft neutral backdrop [style], soft diffused window light from camera left [lighting], 105mm lens, f/2, eye-level, close-up from forehead to chin, sharp focus on the eyes, very soft background [camera], calm, natural tones, natural skin texture with visible pores [mood]. Free from plastic skin, glassy eyes and warped features. Add a small watermark text reading 'www.PromptsForPhotos.Com' in the bottom-right corner, small clean sans-serif font, pure white, 50% opacity, unobtrusive, not overlapping the subject.
5. Top-Down Food Shot (50mm)
Prompt 05
Best for: ChatGPT · Gemini
A wooden table with a white plate of fresh fruit salad, a glass of water and a linen napkin [subject], realistic food photograph in a bright kitchen [style], soft window light from the left [lighting], 50mm lens, f/5.6, top-down bird's-eye view, whole plate in sharp focus [camera], fresh, bright, natural colours [mood]. Free from melted objects, warped text and extra items. Add a small watermark text reading 'www.PromptsForPhotos.Com' in the bottom-right corner, small clean sans-serif font, pure white, 50% opacity, unobtrusive, not overlapping the subject.
Camera Mistakes to Avoid
- Wide lens plus heavy blur. A 24mm lens keeps most of the scene sharp, so it does not match "extreme background blur".
- Very wide lens for close portraits. It stretches faces. Use 50mm to 105mm for portraits.
- Several lenses in one prompt. Choose one.
- No crop named. The AI may give a full-length shot when you wanted a headshot.
- Camera words without light. A great lens look still looks flat if there is no light direction.
Frequently Asked Questions
Do camera and lens words really work in AI image tools?
They work as style signals. The AI is not using a real lens, but it has learned what photos taken with an 85mm lens at f/1.8 usually look like, so the words move the result in that direction. The match is never exact.
What is the best lens for AI portraits?
For flattering portraits, 85mm at f/1.8 to f/2.8 is the most common choice because it gives soft background blur and natural face proportions. For candid, natural shots, 50mm works well.
What does f/1.8 mean in a prompt?
It is a wide aperture setting that gives a shallow depth of field: the subject stays sharp and the background turns into a soft blur. A larger number, such as f/8, keeps more of the scene sharp.
Should I use more than one lens in a prompt?
No. One lens, one aperture and one angle are enough. Several lenses give the AI conflicting instructions, and it may blend them into a vague result.
Why does my wide-angle portrait look distorted?
Very wide lenses stretch faces, especially near the edges of the frame. For close portraits, use 50mm to 105mm. Use wide lenses for scenes and environmental shots.
Can I describe a camera brand or film stock?
You can try it, and some tools respond to words such as '35mm film look'. Results vary by tool, so test it and keep only the phrases that clearly help.
Final Thoughts
Camera words are a small vocabulary with a big effect. Pick a lens that suits the job, choose an aperture for the amount of blur, name one angle and one crop, and send it with clear lighting. Save this page for the tables, test your favourite recipes, and keep the ones that work best in your tool. Our guide to the 15 mistakes that make AI photos look fake shows what to check when a result still looks artificial.
