AI image generators are powerful, but they work with what you give them. A blurry, poorly lit, or oddly framed photo will produce a blurry, poorly lit, or oddly framed result — no matter how good the model is. Choosing the right source photo is the single most impactful thing you can do before you generate.

Lighting: the most important factor

Good lighting reveals the shape of your face. Bad lighting hides it — and the AI will fill in the gaps with guesses that may not look like you.

What to look for

• Even, diffused light — overcast daylight, a window on a cloudy day, or a softbox. Shadows should be soft, not harsh.
• Light facing you — the source should be in front of you, not behind. Backlit photos (where the window is behind you) create silhouettes.
• No mixed color temperatures — avoid mixing warm indoor bulbs with cool daylight. It creates unnatural skin tones that confuse the model.

Framing and composition

The AI needs to see your face clearly. It should occupy a significant portion of the frame — not a tiny speck in a wide landscape shot.

Ideal framing

• Head and shoulders — your face should fill roughly the top half of the image. This gives the model enough detail to work with.
• Face forward or slight angle — a straight-on or three-quarter angle works best. Extreme side profiles lose facial features that the model needs to preserve your likeness.
• Eyes visible — eyes are the most important feature for likeness. Make sure they are clear, open, and not obscured by hair, glasses glare, or shadows.

What to avoid

• Group photos — the model may focus on the wrong person
• Heavy filters or face-altering effects already applied
• Extreme angles (looking straight up or down)
• Sunglasses or anything that hides the eyes
• Motion blur from moving while the photo was taken

Resolution and file quality

Higher resolution gives the model more detail to work with. A photo taken on a modern smartphone is almost always sufficient — the issue is usually compression or cropping, not the camera itself.

• Minimum recommended: 512 × 512 pixels of face area
• Ideal: 1024 × 1024 pixels or larger
• File formats: JPG, PNG, or WEBP — all work equally well
• Avoid: screenshots of screenshots, heavily compressed social media downloads, or photos that have been resized up from a small original

Background

A plain, uncluttered background helps the model focus on you rather than trying to interpret a busy scene. A solid wall, a simple outdoor setting, or even a blurred background all work well.

That said, the AI will replace or stylize the background anyway — so a busy background is not a dealbreaker. It just means the model has to do more work, and the result may be slightly less accurate.

A quick checklist before you upload

1. Face is clearly visible and takes up most of the frame
2. Eyes are open and unobscured
3. Lighting is even — no harsh shadows across the face
4. Photo is in focus — no motion blur
5. No heavy filters or face-altering effects already applied
6. File is at least 512px on the shortest side

Once you have a good source photo, the style you choose determines the final look. Browse the full style library to find the aesthetic that fits you, then head to the Create page to generate your PFP.