TL;DR: This guide gives you 35 tested, copy-paste AI headshot prompts for ChatGPT, Midjourney v7, and Ideogram 3.0 — covering professional, casual, creative, executive, and industry-specific looks. Each prompt has fill-in-the-blank slots for your details. The average AI headshot costs $25–$35 versus $100–$1,200 for a studio session, and recruiters in a 2024 blind test preferred AI headshots 76.5% of the time. Below, you'll find the prompts, the settings that make them photorealistic, the platform-specific tweaks, and the mistakes that turn 80% of AI headshots into unusable rejects.
What are the best AI headshot prompts for ChatGPT, Midjourney, and Ideogram?
The best AI headshot prompts describe four things in order: the subject (age, gender, hair, clothing), the expression, the background, and the lighting — then close with platform settings like --ar 1:1 --raw --v 7 for Midjourney. Front-loading concrete detail and naming a single light source is what separates a usable, photorealistic LinkedIn headshot from a generic, plastic-looking reject.
That single paragraph is the whole game in miniature. Everything else in this guide is depth: the 35 ready-to-use prompts, the reasoning behind each setting, and the workflow tricks that get you a headshot looking like you instead of a stock-photo stranger.
The reason prompting matters so much is economic. The global AI portrait and headshot market generated an estimated $420 million in 2025, up from $180 million in 2022, and professional willingness to use AI headshots jumped from just 8% in 2021 to 58% in 2025. But the same research shows a brutal quality gap: independent reviewers consistently find that only 10–20% of AI-generated headshots are actually usable in a professional context. A good prompt is the difference between landing in that usable 20% on the first batch versus burning credits on faces you'll never post.
How do you use these AI headshot prompts?
Every prompt below has bracketed slots like [gender/age], [hair description], and [clothing]. Replace the brackets with your details, keep the lighting and composition cues, and generate. Unless a prompt says otherwise, it's written for Midjourney v7 — but the descriptive core works in ChatGPT (gpt-image-1) and Ideogram 3.0 too. You just drop the Midjourney -- flags when you move to those tools.
Here's the mental model for each prompt's structure:
| Slot | What it controls | Example |
|---|---|---|
| Subject | Age, gender, build, role | Professional 35yo female |
| Hair | Length, color, style | shoulder-length brown hair |
| Clothing | Garment, color, cut | navy blazer over white blouse |
| Expression | A single emotion | confident slight smile |
| Background | Setting + blur level | neutral grey backdrop |
| Lighting | Source + direction | soft north window light |
| Settings | Ratio, style, version | --ar 1:1 --raw --v 7 |
If you want the headshot to look like your actual face, pair any prompt with a face reference: --cref [URL_to_your_real_photo] in Midjourney, an uploaded reference image in Ideogram, or "make this look like the attached photo" in ChatGPT. This is also the ethical path — more on that below — because AI headshots built from your real likeness are accepted on LinkedIn, while invented faces presented as you are not.
Want these as one-click presets? Prompt Architects ships every prompt in this guide as a saved template with Global Variables, so you fill in your details once and reuse across all three platforms. See our Midjourney prompt library guide for the workflow.
Professional and corporate headshot prompts (10)
These are the workhorses — LinkedIn, company directories, speaker bios, and "About" pages. The common thread is restraint: neutral or lightly blurred backgrounds, one clear expression, and directional light that flatters without theatrics. Keep --raw on every one of these; without it, Midjourney layers on a subtle painterly aesthetic that immediately reads as "not a real photo."
1. Standard LinkedIn — natural light
[Professional 35yo female], [shoulder-length brown hair], [navy blazer over white blouse], confident slight smile, neutral grey backdrop, soft north window light, sharp focus on eyes, professional headshot
--ar 1:1 --s 100 --raw --v 7
2. Executive — moody studio light
[Confident 45yo male executive], [salt-and-pepper hair, neatly groomed], [charcoal suit, white shirt, no tie], serious expression, dramatic single-key light from upper left, deep matte grey background, executive portrait
--ar 1:1 --s 150 --raw --v 7
3. Tech founder — relaxed
[Mid-30s startup founder], [casual button-down shirt rolled at sleeves], approachable smile, slightly out-of-focus modern office in background, golden hour through windows, environmental headshot
--ar 1:1 --s 150 --raw --v 7
4. Investment banker — formal
[Polished 40yo banker], [tailored navy three-piece suit], confident composed expression, dark mahogany office background slightly blurred, side rim light from window, polished traditional finance portrait
--ar 1:1 --s 150 --raw --v 7
5. Doctor — trust-signal portrait
[Female doctor in white coat over scrubs], [warm welcoming expression], stethoscope around neck, soft hospital corridor blurred behind, even diffused light, professional medical headshot
--ar 1:1 --s 100 --raw --v 7
6. Lawyer — authority
[Sharply-dressed lawyer 40s], [pressed navy suit, silk tie], confident direct gaze, leather-bound book wall background heavily blurred, dramatic side light, classic professional portrait
--ar 1:1 --s 150 --raw --v 7
7. Consultant — clean modern
[35yo female management consultant], [sleek bob haircut, minimal makeup], [blazer over knit top], slight smile, clean white seamless backdrop, soft beauty dish lighting, modern professional headshot
--ar 1:1 --s 100 --raw --v 7
8. Realtor — warm and approachable
[Friendly 40s realtor], [warm smile showing teeth], [smart blazer over collared shirt], outdoor location with blurred residential street, golden afternoon light, environmental headshot
--ar 1:1 --s 150 --raw --v 7
9. Startup CEO — magazine cover style
[Mid-40s tech CEO], [crisp white t-shirt], crossed arms, looking off camera with intent, dramatic side light, dark grey background, fortune magazine cover aesthetic
--ar 4:5 --s 200 --raw --v 7
10. Academic / professor
[Distinguished 50s professor], [tweed blazer with elbow patches, button-down], thoughtful expression, slightly blurred bookshelf background, warm desk lamp lighting, intellectual portrait
--ar 1:1 --s 150 --raw --v 7
Why these work: LinkedIn drives an estimated 47% of all headshot demand, and profiles with a professional photo see roughly 21x more profile views and 9x more connection requests than photo-free profiles. The 1:1 ratio is non-negotiable here because LinkedIn crops to a circle inside a square. The --s value climbs with the drama you want — 100 for clean corporate light, 150–200 when you want directional, editorial shadows.
Casual and lifestyle headshot prompts (8)
Lifestyle shots are for dating profiles, personal brands, "meet the team" pages with a human vibe, and creative-industry bios. The rules invert slightly: you want environmental context, candid energy, and golden-hour warmth. Push --s higher (200–250) because a touch of Midjourney's aesthetic actually helps here, and lean toward 4:5 or 3:4 for feed-friendly framing.
11. Dating profile — outdoor cafe
[Mid-30s person], [casual sweater], genuine candid smile, sitting at outdoor cafe, blurred urban background, golden hour light, lifestyle photography
--ar 4:5 --s 200 --raw --v 7
12. Coffee shop natural
[20s creative], [denim jacket over t-shirt], holding ceramic mug, soft window light from left, blurred coffee shop interior, candid lifestyle moment
--ar 4:5 --s 200 --raw --v 7
13. Park lifestyle
[30s], [linen shirt rolled at sleeves], walking in green park, late afternoon light, slight motion blur on background, candid smile, outdoor lifestyle
--ar 3:4 --s 200 --raw --v 7
14. Beach casual
[Mid-30s], [linen shirt unbuttoned, swim trunks], relaxed pose on coastal cliff, ocean horizon behind, golden hour, breezy hair, vacation lifestyle portrait
--ar 4:5 --s 200 --raw --v 7
15. City rooftop
[Late-20s creative], [oversized sweater], leaning on rooftop railing, city skyline blurred behind at twilight, mixed warm and cool light, urban lifestyle
--ar 4:5 --s 250 --raw --v 7
16. Studio casual
[20s], [white tee, denim jacket], hands in pockets, plain warm beige backdrop, soft beauty dish light, casual studio portrait
--ar 4:5 --s 100 --raw --v 7
17. Bookshop interior
[Mid-30s], [knit sweater], browsing bookstore aisle, warm tungsten light, atmospheric depth, candid moment captured
--ar 3:4 --s 200 --raw --v 7
18. Bike commuter
[30s], [cycling cap, casual shirt], pausing on bike at street corner, european city background blurred, late afternoon golden light, urban lifestyle
--ar 4:5 --s 200 --raw --v 7
A note on dating profiles specifically: candid beats posed. The word candid plus genuine in front of smile nudges the model away from the dead-eyed catalog stare that screams "AI." If the result still looks too perfect, add slight imperfection, natural skin texture — at lower stylize values, Midjourney v7 renders texture, imperfection, and natural light falloff without smoothing them, which is exactly what you want for believability.
Creative and artistic portrait prompts (8)
These break the realism rules on purpose — for author photos, band promos, fashion portfolios, and personal-brand work that needs a point of view. Drop --raw (you want the aesthetic now), crank --s way up, and reach for --niji 6 when you want anime-illustrated styling.
19. Editorial fashion
[Tall model 20s], [oversized red wool coat], dramatic geometric pose against cream wall, bold side rim light, fashion editorial photography, high contrast, 85mm
--ar 4:5 --s 250 --v 7
20. Black and white art
[40s artist], [paint-spattered apron over linen shirt], hands holding palette, single overhead window light, deep shadows, ilford hp5 film grain, fine art portraiture
--ar 4:5 --s 200 --v 7
21. Cinematic anamorphic
[Mid-30s], [black turtleneck], side profile lit by single warm practical light, deep shadow on opposite side, anamorphic lens flare, blade runner aesthetic
--ar 21:9 --s 300 --v 7
22. Film noir
[40s detective figure], [trench coat, fedora], silhouetted by venetian blind shadows on face, cigarette smoke curling, 1940s film noir, black and white, high contrast
--ar 4:5 --s 400 --v 7
23. Renaissance painting style
[Distinguished figure 30s], [renaissance robes in deep burgundy], chiaroscuro lighting, dark moody background, vermeer style oil painting, gallery quality
--ar 4:5 --s 600 --v 7
24. Peter Lindbergh inspired
[Effortlessly elegant 40s], [simple white button-down shirt, no makeup], natural beach light, candid laughing moment, peter lindbergh black and white aesthetic, wind-swept hair
--ar 4:5 --s 300 --v 7
25. Studio Ghibli portrait
[Stylized character], [sweater and skirt], window seat with cat, watercolor wash, makoto shinkai style, soft pastel palette, anime portrait
--ar 4:5 --s 750 --niji 6
26. Comic book hero
[Superhero], [costume description], heroic three-quarter pose, dramatic high-contrast lighting, halftone shading, jack kirby comic illustration style
--ar 3:4 --s 600 --v 7
The named-artist references (Lindbergh, Vermeer, Kirby) are stylistic shorthand the model understands. Use them to set a mood, not to copy a living artist's protected work — and never present a stylized illustration as a literal photo of yourself on a professional profile.
Industry-specific headshot prompts (9)
These bundle the costume, props, and environment for a specific profession so you don't have to assemble them yourself. Each is tuned to the trust signals that industry expects — a chef dusted with flour, a guide weathered by the outdoors, a speaker mid-keynote.
27. Wedding officiant / photographer
[Professional wedding photographer], [neutral elegant attire], holding camera, soft outdoor reception light, wedding industry portrait
--ar 1:1 --s 150 --raw --v 7
28. Fitness coach
[Athletic coach 30s], [fitted athletic wear], confident posture, gym environment slightly blurred, dramatic side light, fitness industry portrait
--ar 4:5 --s 200 --raw --v 7
29. Chef in kitchen
[Chef in white uniform], [hands lightly dusted with flour], professional kitchen background, warm overhead lighting, food industry portrait
--ar 4:5 --s 150 --raw --v 7
30. Author / book jacket
[Author in their 50s], [thoughtful expression, comfortable cardigan], home library background heavily blurred, warm lamp light from left, classic literary portrait
--ar 4:5 --s 200 --raw --v 7
31. Musician portrait
[Singer-songwriter 30s], [vintage band tee], holding acoustic guitar, dimly lit recording studio, atmospheric depth, music industry portrait
--ar 4:5 --s 250 --v 7
32. Tech speaker / conference
[Confident speaker 40s], [smart casual outfit], holding lavalier mic, slightly blurred conference stage backdrop, dramatic stage lighting, keynote speaker portrait
--ar 4:5 --s 200 --raw --v 7
33. Outdoor adventure / guide
[Rugged adventure guide], [technical outdoor jacket], mountain landscape behind, golden afternoon light, weathered authentic character, outdoor industry portrait
--ar 4:5 --s 200 --raw --v 7
34. Healthcare professional
[Nurse in scrubs], [warm caring expression], soft clinic background blurred, even diffused lighting, healthcare professional portrait
--ar 1:1 --s 100 --raw --v 7
35. Real estate agent — luxury
[40s realtor], [tailored blazer], standing in front of luxury home exterior at golden hour, confident smile, real estate marketing portrait
--ar 4:5 --s 200 --raw --v 7
One caveat worth knowing: healthcare professionals are the most hesitant group about AI headshots, at 29% adoption, versus 68% for job seekers. Trust signals matter most in medicine, so for a clinical headshot, a photo reference of your real face is strongly advised — the "warm caring expression" only lands if the face beneath it is unmistakably you.
Which AI headshot tool should you use?
Each platform has a clear sweet spot. Here's how to choose:
| Tool | Best for | Realism | Text in image | Speed |
|---|---|---|---|---|
Midjourney v7 + --raw | Photoreal LinkedIn/exec headshots | Highest | Weak | Medium |
| Ideogram 3.0 | Badges, name plates, headshot + text | High | Best in class | Fast |
| ChatGPT (gpt-image-1) | Fast iteration, edits via chat | Good | Good | Fastest |
| Flux / Schnell | Bulk generation on a budget | High | Medium | Very fast |
| HeadshotPro / Aragon AI | Exact-likeness from your photos | High | N/A | Slow (training) |
Midjourney v7 is the realism champion. In standardized testing by one team, v7 produced more photorealistic outputs than v6 in 23 of 30 prompts, with measurable gains in skin texture, fabric detail, and shadow rendering. The catch is the --raw flag: raw mode turns off Midjourney's automatic beautification, which is exactly what a believable headshot needs.
Ideogram 3.0 is the pick when text has to be in the frame — a name plate, a conference badge, a "Realtor of the Year" ribbon. Its 3.0 release pushed hard on image-prompt alignment, photorealism, and text rendering quality, and it consistently scored highest in human ELO evaluations across diverse prompts. Plain headshots still go to Midjourney; headshots with legible typography go to Ideogram.
ChatGPT's image model (gpt-image-1) wins on iteration speed and conversational editing — "make the background warmer," "open his eyes a little" — without re-prompting from scratch. It's weaker on hands and complex multi-source lighting, so use it to nail composition fast, then move the winning concept to Midjourney for the final render.
For a deeper platform-by-platform breakdown, see our comparison of the best AI image generators and our guide to writing prompts that survive across models.
What settings make AI headshots look photorealistic?
Three settings carry most of the realism load in Midjourney, and getting them right is more important than the wording of your prompt.
1. --raw (raw mode). This is the single biggest lever. Raw mode disables the house aesthetic. Without it, even a perfect prompt comes back slightly illustrated. For anyone porting an older prompt library to v7, adding --style raw is the most important adjustment you can make.
2. --s (stylize), kept low. For portraits, a stylize value of 50–100 keeps skin reading as real without looking forensic. At --s 0, the model renders texture and natural light falloff without smoothing — great for grit, sometimes too harsh for corporate. Stay under 200 for any headshot meant to pass as a real photo.
3. Aspect ratio matched to the platform. Use this cheat sheet:
| Platform / use | Aspect ratio | Why |
|---|---|---|
| LinkedIn profile | --ar 1:1 | Crops to a circle in a square |
| Instagram feed | --ar 4:5 | Tallest allowed feed crop |
| Full-body lifestyle | --ar 3:4 | Room for body + context |
| Cinematic / banner | --ar 21:9 | Wide editorial framing |
| Email signature / directory | --ar 1:1 | Universal square cell |
A working reference prompt from Midjourney's own community pattern: a single overhead pendant casting warm shadows below the jawline, shot on a 50mm lens, --ar 3:2 --s 100 --style raw. Notice the discipline — one light source, one lens, a low stylize value. That restraint is the recipe.
How do you make an AI headshot look like your real face?
Text alone gives you a plausible stranger. To get you, anchor the model to a real photo. This also resolves the most common complaint — 63% of dissatisfied users blame insufficient facial likeness.
--cref [URL_to_your_real_photo]in Midjourney carries your face across any of the styles above. Add--cw 100to weight likeness heavily, or lower--cwto let the style bend the face more.--seed [number]locks the subject so you can iterate on outfit and lighting without losing the face you liked.- Upload 1–3 reference images in Ideogram 3.0, which lets you control generations to follow a preferred aesthetic and identity.
- In ChatGPT, attach your photo and say "generate a professional headshot of this person in [style]." Then refine conversationally.
- For perfect likeness with zero prompting, train a LoRA or use a photo-trained service (HeadshotPro, Aragon AI). Expect $25–$100 and a wait while it trains on 10–20 of your selfies.
If you reuse one face constantly, save the reference URL and your favorite prompt as a single reusable template. Our guide to character consistency walks through locking a face across an entire headshot set.
Are AI headshots good enough to replace a photographer?
For most everyday uses, yes — and the data is striking. In a 2024 Ringover survey of 1,087 US recruiters, 76.5% preferred AI-generated headshots over real ones in a blind test, and recruiters correctly identified the AI images only 39.5% of the time — barely better than a coin flip, even though 80% believed they'd spotted them accurately.
But there's a sharp catch on disclosure. The same study found 66% of recruiters would be put off once they knew a headshot was AI-generated, and 88% said AI use should be disclosed. The aesthetic passes; the trust test is where things get delicate.
Here's a practical decision guide:
| Use case | AI headshot verdict |
|---|---|
| LinkedIn, freelance profile, dating | Excellent — indistinguishable from studio work |
| Internal company directory / team page | Excellent — major savings, fast turnaround |
| Conference speaker bio, podcast guest | Very good — pick the editorial prompts |
| Book cover, investor deck, exec PR | Use a photographer or a photo-trained model + disclosure |
| Anything legal/identity-sensitive | Real photography; do not invent a likeness |
The money case is genuinely compelling: the average AI session runs $25–$35 versus $100–$1,200+ for traditional photography, companies populating directories save $2,400–$8,000 a year, and results land in 15–90 minutes versus 3–7 business days. With free prompts and an existing generator subscription, your marginal cost approaches zero.
What are the most common AI headshot prompt mistakes?
These five errors account for most of the unusable 80%:
- Skipping
--rawon a professional headshot. Without it, Midjourney's beautification makes the image read painterly. Always include it for realism. - Stacking conflicting expression cues. "Casual smile" and "serious expression" average into a blank, uncanny stare. Pick exactly one emotion.
- Vague clothing. "Suit" yields a generic suit. "Charcoal three-piece tailored suit, white shirt, navy silk tie" yields the suit you pictured. Specificity is free.
- No lighting cue. Lighting is half the look. Always name a source and direction — "soft north window light," "single overhead key light from upper left."
- Wrong aspect ratio for the platform. A 16:9 headshot gets butchered by LinkedIn's circular crop. Match the ratio to the destination every time.
A sixth, subtler one: over-describing. Twelve adjectives don't beat four precise ones. The model averages everything you give it, so each weak word dilutes the strong ones. Write tight.
How do you build a repeatable headshot workflow?
If you'll generate headshots more than once — for a team, a client roster, or your own quarterly refresh — turn the one-off prompt into a system:
- Lock a base prompt with your face reference and preferred lighting, then change only one variable at a time (outfit, background, ratio).
- Use
--seedto hold the subject steady while you A/B different wardrobes. - Generate batches, not singles. At
--c 15–25you get meaningfully different variants instead of four near-identical frames. - Borrow a look with
--sref [URL]— point it at a headshot whose lighting you love to inherit that mood. - Save the winning combo as a template. This is where Prompt Architects earns its keep: store the prompt with Global Variables (name, role, brand colors), reuse it across Midjourney, Ideogram, and ChatGPT, and share it with your whole team so everyone's directory photo matches.
For teams standardizing a visual identity, pair this with our brand-consistent image prompting guide so every headshot shares the same lighting language and crop.
Frequently asked questions
What is the best AI headshot prompt for LinkedIn?
Specify subject details, a confident slight smile, a neutral or office-blurred background, and soft directional light, then close with --ar 1:1 --raw --v 7. The 1:1 square is essential because LinkedIn crops profile photos to a circle inside a square frame.
Can AI headshots replace professional photographers? For LinkedIn, dating, directories, and freelance profiles, yes — recruiters preferred AI headshots 76.5% of the time in a blind test. For book covers, investor materials, and exec PR, professional photography still wins on authenticity, and you should disclose AI use because 88% of recruiters expect it.
Which AI tool produces the best headshots?
Midjourney v7 with --raw for photorealism, Ideogram 3.0 when you need readable text in frame, ChatGPT for fast iteration, and a photo-trained LoRA or HeadshotPro/Aragon AI for an exact likeness of your own face.
What is the best aspect ratio for a LinkedIn headshot? 1:1 (square). Use 4:5 for social feeds and 3:4 for lifestyle shots. Avoid 16:9 for any headshot.
Should I use --raw on AI headshot prompts?
Yes for photorealism. Raw mode turns off automatic beautification; pair it with --s between 50 and 200, staying under 100 for the most natural skin.
Are AI headshots ethical for professional use? AI headshots built from your own face are widely accepted; invented faces presented as you are not. LinkedIn permits AI photos that reflect your real likeness. Disclose when in doubt — 88% of recruiters say AI use should be revealed.
Why do only some AI headshots look usable? Only 10–20% of raw AI headshots are professionally usable, and 63% of complaints cite poor likeness. Fix it with precise clothing/lighting/expression cues, batch generation, and a face reference.
How much money do AI headshots actually save? $25–$35 per session versus $100–$1,200+ for studios; companies save $2,400–$8,000 a year on directories. 71% of users switch primarily for the cost savings.
By Nafiul Hasan — Founder of Prompt Architects, an AI prompt-engineering platform used to craft and standardize image prompts across Midjourney, Ideogram, ChatGPT, and more. Last updated: June 10, 2026.