See Your Future Kids From Two Photos

The exact prompt to blend you and your partner into AI baby photos. Two photos. One prompt. 30 seconds. Free ChatGPT account, no app, no fee, no waiting list.

What You'll Learn

• The exact copy-paste prompt that blends two faces into realistic kid photos
• What ChatGPT actually does behind the scenes (and what it doesn't)
• The variations to try: different ages, multiple kids, specific traits, family photos
• Fixes for the most common issues

What You'll Need

This works on the free tier of ChatGPT (chat.openai.com). You don't need ChatGPT Plus, you don't need a third-party app, and you don't need to pay anything. The only thing that affects quality is the photos you upload.

• One clear, front-facing photo of you (good lighting, no sunglasses)
• One clear, front-facing photo of your partner (same kind of shot)
• A free ChatGPT account at chat.openai.com
• A reasonable phone or laptop. The AI does all the work

Photo tip: selfies in soft natural light beat flash-lit indoor shots every time. Avoid heavy filters, hats, or anything that hides your facial features. The AI uses what it can see. Give it a clean read on both faces.

The Exact Prompt

Drop both photos into a new ChatGPT chat. Then paste this prompt word for word. Adjust the age and number of kids if you want different outputs.

Generate a realistic photo of what our future child might look like as a 5-year-old, blending features from both faces in the photos I uploaded. Show one boy and one girl side by side. Keep the lighting natural and the background neutral. Be specific about which features came from which parent (eyes, nose, smile, hair color, skin tone) in a short caption underneath each photo.

What ChatGPT Actually Does

1. Reads both faces. The vision model identifies key facial features on both parents: eye color, nose shape, jawline, hair texture, skin tone
2. Predicts genetic blending. It uses common-knowledge genetics rules (dominant vs recessive features, inherited proportions) to estimate which parent's traits would likely show up
3. Generates the image. Image generation creates a realistic 5-year-old face combining the predicted traits. You'll usually get one boy and one girl in the same response
4. Tells you what's from whom. If you ask for it (and we did in the prompt above), it captions which parent each feature came from. This is the eerie part

Variations To Try

Once you have the basic version working, run it again with tweaks. The same two photos can produce wildly different outputs depending on how you prompt.

• Different ages: "as a newborn," "as a 10-year-old," "as a teenager," "as a 30-year-old adult"
• Multiple kids: "show 3 siblings, two boys and one girl" (great for big families)
• Specific traits: "with mom's eyes and dad's smile" to control the blend
• Style: "as a school photo," "as a candid backyard shot," "as a black and white portrait"
• Family photo: "now show all of us as a family photo at the beach in 2035"

Group variation: if you and your partner both have siblings, run the prompt with photos of you + your partner + both your moms. ChatGPT will generate kids with even more reference points, and the results often pull in features from grandparents that wouldn't show up otherwise.

Common Issues

• Output looks generic. Cause: photos too low-resolution or filtered. Fix: use unfiltered, well-lit photos. Avoid Instagram filters
• Skin tone looks off. Cause: AI defaulted to one parent. Fix: add "blend skin tones evenly between both parents" to the prompt
• Eyes look uncanny. Cause: image generation glitch. Fix: just regenerate. Each run is different
• Won't generate a child. Cause: content policy. Fix: reword to "realistic portrait of a fictional 5-year-old descendant"
• Same face for boy and girl. Cause: prompt was too vague. Fix: specify "distinctly different facial structures for the boy and the girl"

Why This Works (And What It Doesn't)

ChatGPT's image generation isn't doing real genetics. It's pattern-matching what a child of two given faces would plausibly look like, based on millions of family photos in its training data. So the output is statistically reasonable but not predictive.

• It captures vibe and family resemblance accurately. Siblings often get this exact look
• It does NOT predict actual traits your kids will have. That's not how genetics or AI works
• Run it 5 times and you'll get 5 different outputs. Pick the one that feels right
• Grandkids work too: "what would my kid's kid look like with this future partner photo"

Don't tell yourself it's real: these images are fun and emotionally surprising. That's the whole point. But don't tell yourself this is what your kids will actually look like. The AI is making an aesthetic guess, not a genetic prediction. Treat it like a Polaroid from a parallel universe, not a forecast.

Go Deeper

Prefer a printable version? Download the PDF: https://drive.google.com/file/d/1dJj7LJGg2IUJPKUm7M9GImU_qiATMz_N/view?usp=drivesdk

Want the whole system installed live, with Cooper on the call? Build Your Own AI, the two week sprint runs the first two weeks of August.

Or get a new working tool every Monday inside The AI Operator community on Whop.