Scene studies / Selfie

AI influencer selfie prompts.

Start with a believable camera position, one light source and something ordinary to do. These five selfie prompts give your AI influencer a different moment without rebuilding the character for every post.

Explore all 5 prompts

Full text · Free to copy · No account needed
Updated October 17, 2026

AI-generated example of soft indoor selfie — Soft indoor selfieAI-generated

Choose your first frame

01Soft indoor selfie 02Outdoor selfie with a dog 03Direct-flash mirror selfie 04Unposed daylight snapshot 05Passenger-seat golden hour

Choose who is holding the camera.

An arm’s-length selfie, a mirror selfie and a photograph taken by someone else are different compositions. Pick one before adding mood words. A phone visible in a mirror makes sense; a phone floating in front of a supposedly front-camera image does not.

These are generated examples from our preset library, using different adult characters. The public prompt uses a neutral 24-year-old subject. Your character, generator, style and crop can change the result. Suggested ratios describe the current preset; example images may use an earlier crop.

Soft indoor selfie

Natural styleSuggested ratio 3:4

Complete prompt

An unpolished, not-professional selfie of a 24-year-old person in a cute oversized hoodie, hair tucked into the hood, soft indoor daylight, relaxed smile

Why this scene works: The hoodie, relaxed expression and soft daylight establish a casual scene without a complicated background or pose.
Make one change: Swap the hoodie color or name a plain wall behind the subject. Keep the hood and soft indoor daylight for the first variation.
Check before publishing: Inspect the hairline where it meets the hood and any lettering the model invents on the garment.

Outdoor selfie with a dog

Natural styleSuggested ratio 3:4

Complete prompt

A casual, slightly elevated selfie of a 24-year-old person with straight hair, posing outdoors nestled close to a large golden retriever with a wavy golden coat, wearing a fitted black cotton top that contrasts with the dog's fur. Blurred green shrubs behind, soft sunlight, dappled leaf shadows on the grass. Informal composition with face and dog centred. Soft even daylight bringing out the golden fur, the black of the shirt and the blue sky through the trees. Shot on a modern smartphone, plenty of detail, moderate digital sharpening, warm and cheerful candid mood.

Why this scene works: The dog gives the subject a reason to lean toward the camera. Greenery and even daylight keep the two faces readable.
Make one change: Try a garden or a shaded park. Keep the animal beside the subject and leave enough room for both heads.
Check before publishing: Check where fur, hands and clothing overlap. Make sure the animal has a coherent face and visible body connection.

Direct-flash mirror selfie

Y2K styleSuggested ratio 9:16

Complete prompt

A 24-year-old person taking a viral Instagram baddie mirror selfie, harsh direct flash, hair styled sleek, confident posture, glossy makeup

Why this scene works: A single flash creates an obvious light direction and the mirror establishes why the phone is in the frame. ‘Viral’ is a style cue in this preset, not a reach prediction.
Make one change: Add a plain dark jacket and a simple hallway mirror. Keep the flash instruction; avoid adding soft window light at the same time.
Check before publishing: Trace the phone and hand through the reflection. Inspect glare, mirror boundaries and the garment before posting.

Unposed daylight snapshot

Smartphone styleSuggested ratio 2:3

Complete prompt

A candid selfie of a 24-year-old person, hair slightly messy, spontaneous snapshot, unposed, natural light

Why this scene works: This short prompt leaves more of the scene open. It is useful for finding a relaxed expression before adding props or a specific location.
Make one change: Add one setting, such as beside a kitchen window. Keep the natural light and unposed direction, then compare with the original.
Check before publishing: Review eye direction and face consistency. A loose prompt gives the generator more room to invent the background.

Passenger-seat golden hour

Camera styleSuggested ratio 2:3

Complete prompt

A relaxed selfie of a 24-year-old person in the passenger seat of a car, seatbelt across a plain knit sweater, hair tousled from the open window. Late-afternoon sun rakes in through the windscreen, warm highlights and soft shadows across the face, blurred trees streaming past outside. Slight handheld tilt, shallow depth of field, natural skin texture, unposed expression mid-laugh, authentic smartphone snapshot.

Why this scene works: The seatbelt, knit sweater and late-afternoon window light anchor the photograph in a recognizable everyday setting.
Make one change: Change the sweater color or describe trees outside a parked car. Keep the subject in the passenger seat and the light coming from one window.
Check before publishing: Follow the seatbelt across the torso and inspect window reflections, upholstery and the hand near the camera.

From one prompt to a repeatable series

Keep the identity. Direct the scene.

Keep the person fixed

Select the same trained character or authorized reference. Match the subject description to that adult character; a new hair color or conflicting face description can pull the result away from the identity you want.

Change one scene detail

Keep the framing and light, then swap the hoodie, room or background. Compare that result before changing the camera angle too. A controlled change tells you which instruction helped.

Review the whole frame

Check the phone, fingers, pupils, mirror edges and background lettering at full size. A convincing face does not make a broken reflection ready to publish.

Troubleshooting

When the result
misses the brief.

The selfie looks like a studio campaign: Keep one everyday light source and an ordinary action. Remove competing instructions such as studio beauty lighting, fashion campaign and flawless skin. Try the candid snapshot as a simpler base.
The mirror shows an extra phone or hand: Specify one person holding one phone in one hand. Simplify the background and choose a chest-up reflection. Review the new image; adding more negative words cannot guarantee correct geometry.
The face changes between posts: Reuse the character/reference and keep identity details stable. Change the scene first. A text description alone does not lock a face across generations.

Before you generate.

How do I write a realistic AI influencer selfie prompt?

Name an adult subject, choose arm’s-length or mirror framing, describe one light source, and add a simple action or expression. Use one of the complete prompts below, then change one scene detail at a time.

How do I keep the same face in different selfies?

Reuse a trained character or a supported identity reference. Keep the subject description consistent with it. The prompt describes the scene; repeating the same words alone does not guarantee the same face.

Do I need a negative prompt for mirror selfies?

These examples use a single text prompt. Start with a simple composition and inspect reflections, hands and phones. If your generator supports a separate negative prompt, use it for a specific defect rather than copying a long unrelated list.

Prompt text is free to copy. Generated-image rights depend on the generator, plan and reference material. In InfluencerForge, commercial work requires a paid plan, subject to the Terms and Usage Policy.

Choose the next scene.

Explore another collection or find these presets in the studio.

Open the preset library

Fitness prompts Travel prompts

Fix a face that drifts between images Train your first consistent character

AI influencer selfie prompts.

Choose who is holding the camera.

Soft indoor selfie

Outdoor selfie with a dog

Direct-flash mirror selfie

Unposed daylight snapshot

Passenger-seat golden hour

Keep the identity. Direct the scene.

Keep the person fixed

Change one scene detail

Review the whole frame

When the resultmisses the brief.

Before you generate.

Choose the next scene.

When the result
misses the brief.