AI Influencer Prompts: 50+ Templates for Nano Banana, Midjourney & Video
A good AI influencer prompt is a photographer's brief: 60 to 120 words covering the subject, outfit, environment, lighting and camera, opened by the same block of character traits every time. Below: a free prompt generator, the prompt formula, a character sheet workflow, prompts for each 2026 model, 50 copy-paste templates and 9 video prompts for Reels.
By Julien — AI & Marketing Expert

Jump to section
Free AI Influencer Prompt Generator
Pick your model, define your character once, choose a niche and a format, and the generator writes a ready-to-paste prompt in that model's style: plain sentences for Nano Banana, scene-only prompts for Higgsfield Soul, parameters for Midjourney, weighted tags and a negative prompt for Stable Diffusion. It runs in your browser, with no sign-up.
AI Influencer Prompt Generator
Build copy-paste prompts optimized for your tool
Choose Your Tool
💡 Nano Banana Pro / Nano Banana 2: attach your character reference image, then paste the prompt. Plain sentences, no parameters or negative prompt.
Define Your Character
These details create your influencer's visual identity
Content Type
Your niche determines locations, outfits, and poses
Scene Setup
Pick or cycle through options — they're niche-matched
Lighting & Camera
Save your character details and reuse them, word for word, in every prompt for a consistent face. The templates below give you 50+ ready-made scenes.
Prefer a distraction-free page? Open the full-page AI influencer prompt generator.
Prompt Anatomy: the 5-Part Brief
Every strong AI influencer prompt answers five questions a photographer would ask: who, wearing what, where, under which light, with which camera. Put your character's fixed traits first, then the five parts, and stop around 60 to 120 words. Longer prompts tend to dilute the details that matter.
[Character invariants] + [Subject & action] + [Outfit] + [Environment] + [Lighting] + [Camera]
| Part | What to specify | Example |
|---|---|---|
| 1. Subject & action | Shot type, pose, expression, one action | "Medium shot, laughing while looking over her shoulder" |
| 2. Outfit | Every garment, color, fabric | "Cream cable-knit sweater, dark straight-leg jeans" |
| 3. Environment | Place type, background details, time of day | "Small café, window seat, rainy street outside" |
| 4. Lighting | Light source and direction | "Soft window light from the left, warm fill" |
| 5. Camera | Lens, aperture, angle, format | "50mm, f/1.8, eye level, vertical 4:5" |
Repeat the invariant traits, word for word
Invariants are the traits that never change: face shape, eye color, hair, skin details, a signature accessory. Write them once as a fixed block and paste it unchanged at the start of every prompt. Rewording it ("dark hair" one day, "brunette" the next) is one of the main reasons a face drifts. In the templates below, this block is the [character] placeholder.
Lena, a 27-year-old woman with shoulder-length dark brown hair and soft curtain bangs, warm olive skin, light freckles across the nose, brown almond-shaped eyes, a small mole above the left eyebrow, thin silver hoop earrings Camera mechanics beat quality adjectives
Words like "hyper-realistic", "8K" or "masterpiece" give the model nothing concrete to render. A lens, an aperture and a light source do. They also push current models away from the glossy, airbrushed look that makes AI images easy to spot.
| Instead of | Write |
|---|---|
| hyper-realistic, 8K, masterpiece | 35mm lens, f/1.8, natural skin texture, visible pores |
| flawless skin, perfect face | light freckles, slight under-eye texture, a few flyaway hairs |
| good lighting | soft window light from the left / direct on-camera flash |
| professional photo | 85mm portrait lens, shallow depth of field, eye-level angle |
| beautiful, stunning | relaxed half-smile, looking just past the camera |
Smartphone-style realism
Most Instagram and TikTok photos are taken on phones, so a "perfect" DSLR look can feel staged. To get the casual phone look, describe the phone's quirks: front camera, wide-angle distortion, everyday background, imperfect framing, flat or mixed light.
Candid smartphone photo of [character] taken with the front camera at arm's length, slight wide-angle distortion, sitting on a bus seat by the window, wearing a gray hoodie, mixed daylight and interior light, slightly off-center framing, natural skin texture, no retouching, vertical 9:16 JSON-style prompts
When you generate in volume, a JSON-style prompt keeps each part of the brief in a labeled field. You swap the outfit or environment field and leave the character field untouched. Natural-language models such as Nano Banana and GPT Image read it as plain text, so it is an organizing method, not a required syntax.
{
"character": "[character]",
"shot": "medium shot, eye level, vertical 4:5",
"action": "laughing while closing a laptop",
"outfit": "oversized white shirt, thin gold necklace",
"environment": "small coworking space, plants, large window",
"lighting": "soft overcast daylight from the window",
"camera": "35mm lens, f/2, natural skin texture",
"avoid": "text, watermark, extra people in focus"
} How to Make a Character Reference Sheet
A character reference sheet is one image of your AI influencer from several angles (front, three-quarter, side) with a few expressions, on a plain background. You attach it to every generation in reference-based tools, and it becomes the visual source of truth next to your text block of invariants.
- Write the invariants block. Face, eyes, hair, skin details, one signature accessory. The character builder helps if you are starting from zero.
- Generate one hero portrait. Iterate until you love it. Drafting with a cheaper model and finishing with a stronger one keeps costs down (for example Nano Banana 2, then Nano Banana Pro).
- Turn the portrait into a sheet with the prompt below, using the portrait as the reference image.
- Build a training set if you need volume. Generate 20 to 30 varied images of the same face (angles, light, outfits). Higgsfield Soul ID trains on 20 to 80 photos; a LoRA needs a similar set.
- Store everything together: the invariants block, the sheet and your 5 best images. Reuse the same files every time.
Using the attached portrait as the only identity reference, create a character reference sheet of the same person on a plain light-gray background. Top row: front view, three-quarter view, side profile and a full-body front view. Bottom row: neutral, smiling and laughing expressions as head-and-shoulders crops. Same face, hair, skin texture and earrings in every panel. Plain white t-shirt and blue jeans, even soft studio light, no text or labels. More methods (face lock, Flux Kontext, LoRA, face swap) in our AI influencer face consistency guide.
AI Influencer Prompts by Model (2026)
The 5-part brief works everywhere, but each model expects it in a different shape. Newer models (Nano Banana, GPT Image, Seedream) want plain sentences and a reference image. Midjourney wants short phrases and parameters. Tools with a trained character (Higgsfield Soul ID, Affogato) want the scene only.
| Model | Prompt style | Same face via | Entry price |
|---|---|---|---|
| Nano Banana Pro / 2 | Plain sentences + edit instructions | Attached reference image | ≈ $0.134/image (API) (November 2026) |
| Higgsfield Soul | Scene only, no face description | Trained Soul ID | $19/mo (November 2026) |
| Seedream 5 | Plain sentences | Reference image, where the platform supports it | Depends on the platform |
| GPT Image | Conversational, detailed instructions | Uploaded reference image | ChatGPT plans or API |
| Midjourney | Short phrases + parameters | Character / omni reference | $10/mo (October 2026) |
| Leonardo AI | Natural language | Character Reference | $12 (October 2026) |
| Kling (video) | Action + camera + duration | Start frame, references | $10/mo (November 2026) |
Prices change often: see our AI influencer tools pricing tracker for current figures.
Nano Banana Prompts (Nano Banana Pro & Nano Banana 2)
Nano Banana Pro (Gemini 3 Pro Image) and Nano Banana 2 (Gemini 3.1 Flash Image) are Google's image models. Both take plain-sentence prompts and can edit an image you attach, which makes them well suited to AI influencers: attach your character sheet, describe the new scene, then refine with short edit instructions. Nano Banana 2 is the default in the Gemini app and suits drafts; Pro is the one for final images. Nano Banana Pro pricing: ≈ $0.134/image (API), as of November 2026. Details in our Nano Banana Pro review.
Formula (with a reference image attached):
Use the attached image as the identity reference: keep the same face, hair, skin texture and proportions. [Shot type] of [character] [action] in [environment]. [Outfit]. Lighting: [light source and direction]. Camera: [lens, aperture, angle]. Natural skin texture, candid photo, no text. Format: vertical 4:5. Example:
Use the attached image as the identity reference: keep the same face, hair, skin texture and proportions. Medium shot of the woman from the reference sitting at a small marble café table, laughing while holding a ceramic cup with both hands. She wears a cream cable-knit sweater and thin silver hoop earrings. Lighting: soft window light from the left, rainy street visible outside. Camera: 50mm lens at f/1.8, eye level. Natural skin texture, candid photo, no text. Format: vertical 4:5. Follow-up edits (one change per message):
Same image, change the sweater to a navy wool coat. Keep the face, pose and lighting exactly the same.
Same image, make it golden hour: warm low sun from the window, longer shadows. Change nothing else.
Same image, switch to a close-up from the shoulders up. Keep the expression and earrings. ▶ Run this prompt in Higgsfield (Nano Banana Pro is included in its plans · affiliate link)
Tips: Stack one change per follow-up: asking for five edits at once is where faces start to drift. No negative prompt needed: write what you want, and add "no text" or "no watermark" if needed.
Higgsfield Soul Prompts (Soul ID)
With Higgsfield, you train a Soul ID on 20 to 80 photos of your character, select it, then prompt the scene. Because the trained character carries the face, your prompt should not re-describe it: extra face details can fight the Soul ID. Focus on action, outfit, setting, light and camera. Plans start at $19/mo (as of November 2026); per-generation costs for Soul are not published, so check the cost shown in the app (see Higgsfield pricing explained).
Formula (Soul ID selected):
[Shot type] of the character [action] in [environment], wearing [outfit]. [Lighting]. [Camera: lens, angle]. Natural skin texture, candid, realistic social media photo. Examples:
Full-body shot of the character crossing a quiet cobblestone street, wearing a beige trench coat, white sneakers and a black crossbody bag. Overcast morning light, 35mm lens at eye level. Natural skin texture, candid, realistic social media photo. Close-up of the character laughing on a sunny balcony, holding an iced coffee, wearing a striped linen shirt. Warm late-afternoon sun from the side, smartphone photo, slight wide-angle distortion. ▶ Run this prompt in Higgsfield (affiliate link)
The same Soul ID carries over to the video models Higgsfield hosts (Kling, Veo, Seedance), so you can animate the images you like without re-describing the character. Our Higgsfield AI review covers the rebuilt AI Influencer Studio, credit rules and refund policy.
Seedream 5 and GPT Image Prompts
Both read prompts the way Nano Banana does: full sentences, the scene described like a brief, and a reference image to hold the face. You usually reach them through ChatGPT or multi-model platforms rather than a dedicated app, so check which reference-image options your platform exposes.
Seedream 5
ByteDance's image model, offered on multi-model platforms such as Framia next to Nano Banana and GPT Image. Write one paragraph, subject first, and keep the invariants block intact.
[character] standing at a rooftop garden bar at sunset, holding a glass of sparkling water with lime, wearing a sage-green silk blouse and wide white trousers. Warm low sun from behind creating a soft rim light, city skyline out of focus. 85mm lens, f/2, eye level, natural skin texture. Vertical 4:5. GPT Image
OpenAI's image model, available in ChatGPT and through the API (and inside platforms such as Leonardo AI, MakeUGC and Framia). In ChatGPT you can describe the scene in plain sentences and ask for changes in the same conversation. If a sign or label must be readable, spell the exact text in quotes and check every output.
Create a photo of the person in the attached image, keeping the same face and hair. She is sitting on the steps of a brownstone, tying the laces of white running shoes, wearing a gray zip-up track jacket and black leggings. Early morning, soft blue light, a coffee cup on the step beside her. Shot like a candid smartphone photo, eye level, vertical 4:5. No text. Midjourney, Flux, Leonardo & Affogato Prompts
These models are still widely used for AI influencers. Midjourney has its own syntax; Flux and Leonardo take natural language; Affogato only needs the scene once your face is locked.
Midjourney
Short, comma-separated descriptions plus parameters. Keep the stylize value low for realism. For the same face, use a character reference (--cref, V6) or Omni Reference (--oref, V7); check the docs for the version you run. Plans start at $10/mo (as of October 2026).
| Parameter | Use |
|---|---|
| --ar 4:5 / 9:16 | Feed / Stories and Reels |
| --style raw | Less beautification, more realistic |
| --s 50–150 | Lower stylize = more literal |
| --cref / --oref [URL] | Same face from a reference image |
| --no text, watermark | Exclude elements |
27-year-old woman with shoulder-length dark brown hair and curtain bangs, light freckles, brown almond eyes, thin silver hoops, laughing at a café window table, cream cable-knit sweater, soft window light from the left, 50mm lens f/1.8, natural skin texture, candid lifestyle photo --ar 4:5 --style raw --s 80 Flux
Full sentences, most important element first, no keyword lists. For edits on an existing image (new outfit, same face), use Flux Kontext and phrase the prompt as an instruction.
A medium shot of [character] sitting in a sunlit café by the window, laughing naturally while holding a ceramic coffee cup. She wears a cream knit sweater. Soft morning light creates gentle shadows on her face, with visible skin texture. Shot on a 50mm lens at f/1.8, shallow depth of field, warm background bokeh. Leonardo AI
Beginner-friendly, with several models in one account (including Nano Banana Pro and GPT-Image-1) and a Character Reference option. Natural-language prompts work; a short negative prompt helps on some models. Plans start at $12 (as of October 2026). Read our Leonardo AI review.
Medium shot of [character] in a sunlit café, cream knit sweater, warm window light with gentle shadows, 85mm portrait lens, shallow depth of field, candid expression, natural skin texture, lifestyle photography Affogato (formerly RenderNet)
Lock a face once, then reuse it: your prompt covers the scene and styling only. Plans start at $24/mo (as of November 2026). Read our Affogato review.
wearing a cream knit sweater, sitting in a sunlit café, laughing candidly while holding a coffee cup, warm window light, 50mm lens, shallow depth of field, natural skin texture, lifestyle photography 50+ Copy-Paste AI Influencer Prompt Templates
50 scene templates in 8 categories, plus 9 model-specific video prompts below. Each template is a complete scene. Replace [character] with your invariants block (or "the person in the reference image" for Nano Banana, GPT Image and Seedream; leave it as "the character" for Higgsfield Soul). Replace [product] with a plain description. The format is written in words, so the templates work across models; for Midjourney, swap "Vertical 4:5" for --ar 4:5.
Instagram feed (4:5)
Your grid posts. Vertical 4:5 fills the most screen space in the feed.
#1 Café window portrait
[character] sitting by a café window with a cappuccino, looking slightly off-camera, wearing a camel turtleneck. Soft window light from the left, 50mm lens at f/2, shallow depth of field, natural skin texture. Vertical 4:5. #2 Golden-hour street portrait
Medium shot of [character] on a quiet tree-lined street at golden hour, wearing a white button-up shirt with rolled sleeves, warm backlight in the hair, candid half-smile. 85mm lens at f/1.8. Vertical 4:5. #3 Flower market
[character] holding a bouquet of tulips at a weekend flower market, wearing a striped long-sleeve top and jeans. Overcast daylight, 35mm lens, documentary style, slightly muted colors. Vertical 4:5. #4 Bookstore aisle
[character] browsing an independent bookstore and pulling a hardcover from the shelf, wearing a green cardigan. Warm tungsten light, 35mm lens at f/2, shot from the end of the aisle. Vertical 4:5. #5 Rooftop at blue hour
[character] leaning on a rooftop railing at blue hour, city lights starting to glow behind, wearing a black leather jacket. Light from a nearby string of bulbs, 50mm lens at f/1.8. Vertical 4:5. #6 Plain-wall close-up (carousel slide 1)
Close-up portrait of [character] against a plain warm-beige wall, relaxed expression, small gold earrings visible. Soft diffused daylight, 85mm lens, visible skin texture and a few flyaway hairs. Vertical 4:5. #7 Candlelit dinner
[character] laughing at a candlelit dinner table, friends partly visible and out of focus at the edges of the frame, wearing a dark green blouse. Warm candlelight, 35mm lens, slight motion blur. Vertical 4:5. Stories & Reels covers (9:16)
Phone-style shots that look native in Stories. Keep key details away from the top and bottom edges: the Stories interface covers them.
#8 Hallway mirror photo
Phone mirror photo of [character] in a hallway mirror, phone held at chest height, wearing a denim jacket over a white t-shirt. Overhead hallway light, slightly tilted framing, everyday clutter in the background. Vertical 9:16. #9 Morning coffee POV
[character] making pour-over coffee in a small apartment kitchen, shot from across the counter at eye level. Morning light through blinds casting stripes on the wall, smartphone photo. Vertical 9:16. #10 Q&A background selfie
Front-camera selfie of [character] at arm's length, sitting on a sofa with a throw blanket. Warm lamp light, slight wide-angle distortion, room in soft focus, empty space at the top of the frame for a question sticker. Vertical 9:16. #11 Outfit decision
[character] sitting cross-legged on the floor surrounded by outfit options, holding up two shirts to the camera with a questioning look. Bright daylight from a window, candid phone photo. Vertical 9:16. #12 Subway commute
[character] standing in a subway car holding the overhead rail, earbuds in, wearing a beige trench coat. Fluorescent light mixed with daylight, candid documentary feel, shot on a phone. Vertical 9:16. #13 Rainy neon street
[character] walking under neon signs on a rainy evening street, holding a clear umbrella, wearing a long dark coat. Reflections on wet pavement, direct on-camera flash. Vertical 9:16. Lifestyle
Everyday scenes that make the persona feel real. These carry most of a lifestyle account.
#14 Home desk
[character] working at a tidy home desk with a laptop and an open notebook, a plant on the windowsill, wearing a soft gray sweater. Natural daylight, 35mm lens at f/2.8, calm focused expression. #15 Plant care
[character] watering houseplants in a sunlit living room, wearing linen overalls over a white t-shirt. Morning sun, 35mm lens, warm and airy color palette. #16 Baking at home
[character] rolling out dough on a floured wooden counter, sleeves pushed up, a smudge of flour on one cheek. Warm kitchen light, 50mm lens, candid moment. #17 Canal bike ride
[character] riding a vintage bicycle with a basket of groceries along a canal path, wearing a striped sweater. Late-afternoon sun, 85mm lens panning shot, background slightly motion-blurred. #18 Park picnic
[character] sitting on a picnic blanket in a park reading a paperback, wearing a sundress with a light cardigan. Dappled sunlight through trees, 50mm lens at f/2. #19 Record shop
[character] flipping through records in a vinyl shop, wearing a corduroy jacket. Warm overhead light, 35mm film look with soft grain. #20 Rainy window seat
[character] curled up on a window seat with a mug of tea, rain on the glass, wearing a chunky knit sweater. Soft gray daylight, 50mm lens, quiet cozy mood. Fitness
Training content reads as real when the form looks right and the clothes look worn. Describe one movement, not a whole workout.
#21 Sunrise trail run
[character] running on a forest trail at sunrise, mid-stride, wearing a long-sleeve running top and running tights. 200mm telephoto compression, soft backlight through the trees, sharp focus on the face. #22 Kettlebell swing
[character] doing a kettlebell swing in a bright industrial gym, wearing a fitted t-shirt and training shorts, focused expression. Side view, 35mm lens, natural light from large windows. #23 Lakeside yoga
[character] holding a tree pose on a yoga mat on a wooden deck by a lake, wearing a long-sleeve fitted top and leggings. Early-morning mist, 50mm lens, calm expression. #24 Post-workout bench
[character] sitting on a gym bench after a workout, towel around the neck, drinking from a water bottle, wearing a zip-up training jacket. Overhead gym lights, candid phone photo. #25 Pilates studio
[character] on a reformer in a bright Pilates studio, wearing a matching long-sleeve set. Side angle, soft daylight, clean minimal background, 35mm lens. #26 Meal prep
[character] packing meal-prep containers of rice, vegetables and chicken on a kitchen counter, wearing a hoodie. Bright daylight, three-quarter overhead angle. Travel
Name the place type, not a famous landmark: generic settings avoid errors and keep your persona the focus.
#27 Airport gate
[character] sitting by an airport gate window with a carry-on suitcase, planes visible outside, wearing a cozy travel set and white sneakers. Flat daylight, 35mm lens, candid. #28 Hill-town steps
[character] climbing stone steps in a Mediterranean hill town with whitewashed walls and blue doors, wearing a loose linen shirt and wide trousers. Harsh midday sun with crisp shadows, 24mm lens. #29 Train window
[character] looking out of a train window at passing mountains, reflection faintly visible in the glass, wearing a knit beanie. Soft overcast light, 50mm lens. #30 Night market
[character] tasting street food at a busy night market, steam rising from the stalls, wearing a light jacket. Warm string lights and mixed neon, 35mm lens, documentary style. #31 Shoreline at sunset
[character] walking along the shoreline at sunset, sandals in one hand, wearing a white linen shirt dress, wind in the hair. 85mm lens, warm backlight. #32 Harbor-view breakfast
[character] having breakfast on a hotel balcony overlooking a harbor, coffee and croissants on the table, wearing a short-sleeve button-up shirt. Morning sun, 35mm lens. Fashion
List every garment, color and fabric. Vague outfits ("stylish look") change from one image to the next.
#33 Concrete-wall OOTD
Full-body shot of [character] standing against a textured concrete wall, hands in pockets, wearing a charcoal oversized blazer, white t-shirt, straight-leg jeans and black loafers. Overcast daylight, 50mm lens, street-style photography. Vertical 4:5. #34 Crosswalk stride
[character] mid-stride on a city crosswalk, wearing a camel wool coat, black turtleneck and ankle boots, the coat moving with the step. 85mm lens from across the street, cool winter light. #35 Accessories detail
Close-up of [character]'s hands adjusting a silver watch and stacked rings, the sleeve of a navy knit visible. Soft window light, 100mm macro lens, shallow depth of field. #36 Autumn capsule
[character] sitting on a park bench among fallen leaves, wearing a rust-colored knit sweater, plaid scarf and brown boots. Warm low sun, 85mm lens at f/2, editorial but relaxed. #37 Studio lookbook
Full-body studio shot of [character] on a seamless light-gray backdrop, one step forward, wearing a pleated midi skirt and a fitted knit top. Softbox key light from the left, 70mm lens, clean lookbook style. #38 Flash editorial
[character] leaning against a vintage car at night, wearing a leather jacket and wide-leg trousers. Direct on-camera flash, dark background, 35mm point-and-shoot look with film grain. Brand content
Replace [product] with a plain description, then composite the real packaging or screen in an editor: models still garble labels and logos.
#39 Skincare hold
[character] holding a frosted glass serum bottle next to the cheek, label facing the camera, bathroom shelf in soft focus. Bright diffused morning light, 85mm lens, clean commercial look, product in sharp focus. #40 Coffee to go
[character] taking the first sip from a takeaway cup outside a café, cup held at chest height with the logo side facing the camera. Morning light, 50mm lens, candid moment. #41 Desk unboxing
[character] unboxing [product] at a clean desk, lid lifted halfway, genuinely curious expression. Soft top light, 35mm lens, product clearly visible and in focus. #42 Fitting-room try-on
[character] in a fitting-room mirror trying on [product], price tag still on the sleeve. Warm boutique lighting, phone photo. Vertical 9:16. #43 Drink pour
[character] pouring [product] into a glass on a sunny kitchen island, label facing the camera, condensation on the bottle. Bright natural light, 50mm lens. #44 App on screen
[character] on a sofa holding a phone with the screen angled toward the camera, showing [app screen], relaxed natural expression. Warm evening lamp light, 35mm lens. Video (image-to-video)
Start from an approved still of your character so the face is already right. One main action per clip, plus a camera instruction and a duration.
#45 Outfit reveal
The character steps into frame from the right, stops in the center and does a slow half turn to show the outfit. Static camera at hip height, daylight, 5 seconds, no cuts. #46 Latte tilt-up
Close-up of the character's hands pouring milk into a latte, then the camera tilts up to the face as the character smiles at the lens. Warm café light, 6 seconds. #47 Walk and talk
The character walks along a sunny sidewalk talking to the camera with natural hand gestures. Handheld camera moving backward at walking pace, light street ambience, 8 seconds. #48 Product lift
The character lifts [product] from below the frame into the light and holds it beside the face, label toward the camera. Locked-off camera, softbox light, 5 seconds. #49 Viewpoint reveal
Wide shot of the character standing at a mountain viewpoint, back to the camera. The camera slowly cranes up to reveal the valley. Golden hour, 6 seconds. #50 Squat set
Side view of the character doing three slow bodyweight squats with steady breathing. Static camera at waist height, bright gym light, 6 seconds. Christmas & Holiday Photoshoot Prompts
A holiday shoot works when only the season changes: same face, same hair, new sweater, new light. These 10 AI Christmas photo prompts cover the December posts most AI influencers need, from the tree and the Christmas market to New Year's Eve. The Nano Banana prompts start from your reference image; the Higgsfield Soul prompts assume your Soul ID character is selected. For 35 more, including portraits from your own photo, couples, product shots and Christmas cards, see our AI Christmas photo prompts.
- •Let the light do the work. Fairy lights, firelight and snowy daylight say "holidays" more than props piled into the frame.
- •Keep it generic. No brand logos, no trademarked mascots or film characters: models often refuse them, and they are an IP risk on sponsored posts.
- •Batch the season. Generate the whole set in one session with the same reference so every post matches. The prompt generator has a Holiday preset if you want variations.
Tree-lights living room Nano Banana
Use the attached image as the identity reference: keep the same face, hair, skin texture and proportions. Medium shot of the woman from the reference sitting cross-legged on a wool rug beside a decorated Christmas tree, untangling a string of warm fairy lights, wearing a cream cable-knit sweater and red plaid pajama pants. Lighting: warm glow from the tree lights and a table lamp, soft shadows. Camera: 35mm lens at f/2, eye level. Natural skin texture, candid photo, no text. Format: vertical 4:5. Hot chocolate close-up Higgsfield Soul
Close-up of the character holding a mug of hot chocolate with marshmallows in both hands, smiling at the camera, wearing a knitted holiday sweater with a snowflake pattern. Warm kitchen light, string lights out of focus behind her. Smartphone photo, slight wide-angle distortion, natural skin texture. Christmas market walk Higgsfield Soul
Full-body shot of the character walking through an outdoor Christmas market at dusk, wearing a long camel wool coat, a burgundy scarf and leather gloves, wooden stalls with warm lights behind her. Light snowfall, 50mm lens at eye level, shallow depth of field. Natural skin texture, candid, realistic social media photo. Gift-wrapping table Nano Banana
Use the attached image as the identity reference: keep the same face, hair, skin texture and proportions. High-angle medium shot of the woman from the reference at a wooden table, wrapping a box in brown kraft paper and red ribbon, pine sprigs, scissors and a roll of twine around her, wearing a soft sage-green sweater. Lighting: soft daylight from a window on the left. Camera: 35mm lens, slightly above eye level. Natural skin texture, candid photo, no text, no logos. Format: vertical 4:5. Gingerbread baking Higgsfield Soul
Medium shot of the character in a cozy kitchen decorating gingerbread cookies with white icing, a light dusting of flour on her apron, wearing a red knit sweater with rolled-up sleeves. Soft window light from the left, 35mm lens at eye level. Natural skin texture, candid, realistic social media photo. Fireplace reading Nano Banana
Use the attached image as the identity reference: keep the same face, hair, skin texture and proportions. Medium shot of the woman from the reference curled up in an armchair by a stone fireplace, wrapped in a chunky knit blanket, reading a hardcover book, wearing a gray turtleneck and wool socks. Lighting: warm firelight from the right, dim room. Camera: 85mm lens at f/1.8, eye level. Natural skin texture, candid photo, no text. Format: vertical 4:5. Snowfall portrait Higgsfield Soul
Close-up portrait of the character outdoors in light snowfall, snowflakes on her knit beanie and eyelashes, wearing a white puffer jacket, cheeks slightly flushed from the cold, soft smile. Overcast winter daylight, 85mm lens at f/2, snowy pine trees out of focus behind her. Natural skin texture, realistic photo. Christmas morning Nano Banana
Use the attached image as the identity reference: keep the same face, hair, skin texture and proportions. Medium shot of the woman from the reference on a sofa on Christmas morning, opening a small gift box and laughing, wearing red-and-white striped flannel pajamas and fuzzy socks, a mug of coffee on the side table and a decorated tree behind her. Lighting: soft morning window light plus the tree lights. Camera: 35mm lens at f/2.8, eye level. Natural skin texture, candid photo, no text. Format: vertical 4:5. Holiday outfit mirror selfie Higgsfield Soul
Mirror selfie of the character in a hallway decorated with a pine garland and warm string lights, phone held at chest height so the face stays visible, wearing a dark green satin blouse, high-waisted black trousers and gold hoop earrings. Warm indoor light, smartphone photo, slight wide-angle distortion, natural skin texture. New Year's Eve rooftop Nano Banana
Use the attached image as the identity reference: keep the same face, hair, skin texture and proportions. Medium shot of the woman from the reference on a rooftop terrace on New Year's Eve, holding a lit sparkler and smiling at the camera, wearing a black velvet long-sleeve midi dress with a tailored wool coat over her shoulders, city lights out of focus behind her. Lighting: sparkler glow on her face and ambient city light. Camera: 50mm lens at f/1.8, eye level. Natural skin texture, candid photo, no text. Format: vertical 4:5. Video Prompts for AI Influencer Reels
A video prompt adds three things to the image brief: one main action, one camera movement and a duration. Start from an approved still of your character (image-to-video) so the face is already right, keep the clip to a single beat, and put spoken lines in quotes only on models with native audio.
[Start frame] + [Action] + [Camera movement] + [Light & setting] + [Duration & format] + [Sound or dialogue]
- Kling: Kling 4.0 (available to everyone since October 15, 2026) makes clips of 3 to 30 seconds, up to 4K and 21:9, with stereo audio, up to 10 keyframes and up to 15 references. Kling 3.0 adds multi-shot generation and native lip-sync in English, Chinese, Japanese, Korean and Spanish. Plans from $10/mo (as of November 2026).
- Seedance: ByteDance's video models for text-to-video and image-to-video. Seedance 2.5 is offered on platforms such as Framia, Pika and MakeUGC, and Higgsfield hosts Seedance too. Its official app is ByteDance's Dreamina, from $15/mo (as of October 2026).
- Veo 3.1: Google's video model with synced audio, available through platforms such as Framia and MakeUGC (Higgsfield and Leonardo AI also host Veo models). Good for short talking clips.
Kling 4.0 · 10 s · 9:16
Image-to-video from the approved start frame. The character opens the curtains, sunlight floods the room, then turns to the camera and stretches. Camera: slow dolly-in from the doorway at eye level. Soft morning light, quiet room tone and birdsong. 10 seconds, vertical 9:16, no cuts. Kling 4.0 (keyframes) · 15 s · 9:16
Keyframe 1 (0 s): the character sits at a café table typing on a laptop. Keyframe 2 (7 s): the character closes the laptop and picks up the coffee cup. Keyframe 3 (15 s): the character stands and walks out of frame to the left. Static medium shot, natural daylight, café ambience. Vertical 9:16. Kling 4.0 · 8 s · 21:9 (YouTube intro)
Cinematic wide shot of the character walking along a cliff path above the sea, camera tracking alongside at a distance, wind moving the hair and jacket. Golden-hour light, ocean and wind sounds. 8 seconds, 21:9. Kling 3.0 (lip-sync) · 8 s · 9:16
The character sits at a desk facing the camera and says: "Three things I wish I knew before my first trip to Lisbon." Natural head movement and small hand gestures, static medium close-up, soft key light. 8 seconds, vertical 9:16. Kling 3.0 (multi-shot) · 3 shots · 9:16
Shot 1: wide shot of the character entering a small bakery. Shot 2: close-up of the character's hands choosing a croissant. Shot 3: medium shot of the character taking a bite and smiling at the camera. Warm interior light, same outfit in every shot (green coat, cream scarf). Vertical 9:16. Seedance · 5 s · 9:16
The character spins once in a flowing midi skirt on a rooftop at sunset, skirt and hair moving naturally. Static camera at chest height, warm backlight. 5 seconds, vertical 9:16. Seedance · 8 s · 9:16
Handheld follow shot behind the character walking through a busy flower market; the character turns and looks back over the shoulder at the camera. Natural daylight, light camera shake. 8 seconds, vertical 9:16. Veo 3.1 (dialogue) · 8 s · 9:16
Medium close-up of the character in a sunny kitchen holding a green smoothie, saying to the camera in a friendly tone: "This is my go-to breakfast when I'm short on time." A blender hum fades out in the background. Natural daylight, handheld phone-camera feel. 8 seconds, vertical 9:16. Veo 3.1 (ambient) · 8 s
The character jogs up a flight of outdoor stairs at dawn toward a static camera at the top, stops and catches their breath while looking at the skyline. Footsteps and soft city ambience, cool blue light turning gold. 8 seconds. Posting on TikTok? See our AI influencer TikTok guide for formats and posting cadence.
Legacy Models: Stable Diffusion 1.5 and SDXL
SD 1.5 and SDXL still make sense if you run models locally with a LoRA of your character. They need a different syntax: comma-separated tags, weights such as (natural skin texture:1.2), and a separate negative prompt. Newer natural-language models need neither. Read our Stable Diffusion review.
Prompt (SDXL):
photo of a 27-year-old woman, shoulder-length dark brown hair, curtain bangs, brown almond eyes, light freckles, (natural skin texture, visible pores:1.2), cream knit sweater, sitting in a sunlit café, warm window light, 50mm lens, shallow depth of field, candid expression, lifestyle photography Negative prompt:
(worst quality, low quality:1.4), bad anatomy, bad hands, extra fingers, missing fingers, deformed, blurry, watermark, text, logo, cartoon, 3d render, anime, illustration, plastic skin, oversaturated, cross-eyed, asymmetric eyes Common AI Influencer Prompt Mistakes
- Rewording the character: "brunette" in one prompt, "dark hair" in the next. Paste the same invariants block every time.
- Quality keywords instead of camera details: "8K, masterpiece" → "50mm, f/1.8, soft window light, natural skin texture".
- Describing the face on Soul ID or face-lock tools: the trained character already carries it, and extra face details can fight it.
- Five edits in one message: with Nano Banana and GPT Image, change one thing per follow-up.
- Cropping after generation: generate in the final format (4:5 feed, 9:16 Stories and Reels, 16:9 YouTube).
- Too much happening in one video clip: one action and one camera move per clip; chain clips in an editor.
- Vague outfits: "stylish outfit" changes every time. Name each garment and color.
Frequently Asked Questions
- What is the best prompt structure for an AI influencer?
- Write it like a photographer's brief of about 60 to 120 words with five parts: subject and action, outfit, environment, lighting, and camera (lens, aperture, shot type). Start every prompt with the same block of invariant traits (face, hair, skin details, signature accessories) so your character stays recognizable from one image to the next.
- How do I make AI influencer images look realistic?
- Replace vague words like "hyper-realistic" or "8K" with camera mechanics: a lens (35mm, 85mm), an aperture (f/1.8), a light source (window light, direct flash) and "natural skin texture". For a social-media look, describe a phone photo: front camera, slight wide-angle distortion, imperfect framing and everyday background details.
- How do I keep the same face in every AI influencer prompt?
- Combine three things: the same invariant-traits block in every prompt, a character reference sheet (front, three-quarter, side and expressions) attached to every generation in tools such as Nano Banana Pro, and a tool-level consistency feature for volume, such as Higgsfield Soul ID (trained on 20 to 80 photos), Midjourney's character or omni reference, or a LoRA. Our face consistency guide compares these methods.
- Do Nano Banana prompts work differently from Midjourney prompts?
- Yes. Nano Banana Pro and Nano Banana 2 read plain sentences and follow edit instructions on an image you attach, so you describe the scene like a brief and refine it with one change at a time. Midjourney uses short comma-separated descriptions plus parameters such as --ar for the aspect ratio, --style raw for realism and --no for exclusions.
- Is there a free AI influencer prompt generator?
- Yes. Our free AI influencer prompt generator builds a ready-to-paste prompt from your character, niche, format, lighting and camera choices, in the right format for Nano Banana, Higgsfield Soul, Midjourney, Flux, Stable Diffusion, Leonardo AI or Affogato. It runs in your browser with no sign-up, on this page and at /toolkit/prompt-generator.
- How do I write video prompts for AI influencer Reels?
- Start from an approved still of your character (image-to-video), then describe one main action, one camera movement (static, push-in, tracking, orbit), the light, the duration and the format (9:16 for Reels). Add spoken lines in quotes only on models with native audio, such as Kling 3.0 and later or Veo 3.1.
- How do I make Christmas photos of my AI influencer?
- Keep the character identical and change only the season: attach your reference image in Nano Banana Pro (or select your Soul ID on Higgsfield), then describe a holiday scene with a seasonal outfit and a light source such as fairy lights, firelight or snowy daylight. Avoid brand logos and trademarked characters. This guide has 10 ready-to-paste Christmas and New Year prompts, and the free prompt generator has a Holiday preset.
- Should I use JSON prompts for an AI influencer?
- They help when you generate in volume. A JSON-style prompt puts each part of the brief (character, outfit, environment, lighting, camera) in a labeled field, so you change one field and keep the rest identical. Natural-language models such as Nano Banana and GPT Image read it as text; it is a way to organize the prompt, not a required format.
What to Read Next
- Face Consistency Guide — Same face across every image
- How to Create an AI Influencer — Step-by-step from concept to first post
- AI Influencer on Instagram — Formats, posting and growth
- Best AI Influencer Generators — Compare tools and pricing