Before and after examples
Portrait prompt
AI Main Character Street Selfie Prompt for ChatGPT AI Images
Copy a curated AI Main Character Street Selfie Prompt for ChatGPT image generation with examples, negative prompt and practical tips.
Copy this prompt
Use the curated prompt as a starting point, then customize the subject, style, lighting and mood in your image tool.
Best for
ChatGPTPortraitReference photo
Improved prompt
Transform the uploaded selfie/portrait into a cinematic "main character street selfie". Keep the person clearly recognizable and preserve their natural facial features, hairstyle, expression, clothing style, and identity. Create a single clean photo only, not a contact sheet, not a collage, not a mockup, not a file preview. No text, no labels, no UI elements, no filenames, no borders. Do not crop the head, face, neck, shoulders, or torso awkwardly. Keep natural upper-body selfie framing with the full face clearly visible. Keep hands and arms natural; avoid broken fingers or warped selfie arms. Make it look like a spontaneous phone selfie taken during a stylish city walk at golden hour or blue hour. Completely replace non-city backgrounds with a modern urban street scene. Add realistic street depth, soft city lights, blurred pedestrians, warm reflections in shop windows, subtle lens flare, natural skin texture, light wind in hair or clothes, and a confident main-character mood. The background should feel alive but not distracting: cafes, crosswalks, city buildings, neon signs, passing cars, or rainy pavement reflections. Make the image look authentic, modern, expensive, and emotionally cinematic, not like a studio portrait or obvious AI edit.
Bad prompt vs improved prompt
Bad prompt
Turn this photo into a cool city selfie.
Improved prompt
How to use this prompt
- Upload a clear selfie or portrait where the face, hair and clothing are visible.
- Copy the full prompt and keep the identity-preserving and framing requirements.
- Choose golden hour, blue hour, rainy pavement or neon street details to control the mood.
- Keep the negative prompt when the output shows labels, UI elements, cropped faces or warped selfie arms.
- Refine only one detail at a time, such as lighting, background depth, crop or facial consistency.
Best photo types for this prompt
Use source images that make the main subject easy to understand. These input types usually give the model enough visual context without adding unnecessary clutter.
Clear selfies with visible facial featuresWaist-up portraitsCasual outfit photosTravel or lifestyle portraitsSocial media profile photos
Best settings for this AI image prompt
Aspect ratios4:5 for social portraits, 3:4 for vertical selfie framing, 1:1 for profile-style crops
Best modelsChatGPT image generation, Midjourney, Flux, Leonardo, Ideogram
Style strengthUse medium style strength for a believable phone selfie and medium-high only when the city lighting needs a stronger cinematic mood.
Input tips
- Use a clear selfie or portrait with the full face visible.
- Keep identity, hairstyle, expression and clothing instructions in the prompt.
- Ask for natural upper-body framing to avoid awkward crops.
- Use golden hour, blue hour or rainy reflections for stronger atmosphere.
- Use the negative prompt if the result shows labels, UI elements, file-preview artifacts or a changed face.
Negative prompt
No distorted face, no changed identity, no plastic skin, no over-retouching, no fake smile, no extra fingers, no warped phone, no duplicated person, no unreadable text, no fake logos, no celebrity likeness, no cartoon style, no extreme beauty filter, no unrealistic lighting, no messy background, no glossy AI look, no contact sheet, no collage, no labels, no file preview, no cropped head, no cut-off face, no awkwardly cut torso.
Common mistakes
- Using a blurry selfie or heavily filtered portrait.
- Removing the instruction to keep the person recognizable.
- Forgetting the natural upper-body selfie framing requirements.
- Making the image too polished like a studio fashion campaign.
- Adding too many signs, logos or background text.
- Skipping negative prompt terms for cropped faces, file-preview artifacts, warped hands or changed identity.
FAQ
What is this main character street selfie prompt used for?
Use it to turn a portrait or selfie into a cinematic city-walk image while keeping the person recognizable. It works best when you want a modern phone-photo look with street lights, natural motion and editorial mood.
Can I use this prompt in ChatGPT?
Yes. The prompt is written for ChatGPT image generation and can be copied directly with an uploaded selfie or portrait.
How do I keep the face recognizable?
Use a clear source portrait and keep the identity-preserving instructions in the prompt. Avoid heavy filters, sunglasses, extreme crops and very low-resolution images.
What kind of photo works best?
A clear selfie, waist-up portrait or casual outfit photo works best. The face, hair and clothing should be visible enough for the model to preserve them.
How do I avoid cropped faces or awkward selfie arms?
Keep the framing requirements in the prompt and use the negative prompt terms for cropped head, cut-off face, awkward torso, broken fingers and warped selfie arms.
Can I change the city background?
Yes. Replace the background details with a rainy street, neon district, cafe sidewalk, crosswalk, city window reflection or another believable urban setting.
Can I use this in Midjourney or Flux?
Yes. Keep the same visual instructions and adapt the aspect ratio or image-reference settings for the tool you use.
What negative prompt should I use?
Use negative terms for changed identity, distorted face, plastic skin, unreadable text, fake logos, collage layout, file preview, cropped head, extra fingers and warped phone details.