Name every input.
Say that each portrait is an identity reference. Without that, a generator tends to invent extra people or blend two faces into one.
Free · private · no photo upload
Choose the setting, feeling, and frame. Get a specific group-photo prompt you can copy for any image tool—or skip prompt-writing and use a ready-made TogetherLens scene.
02 / COPY OR SHARE
Rooftop · natural · half heightCreate one natural group photo using each supplied portrait as an identity reference. Place the group on a simple city rooftop during warm golden hour. Use a warm, candid mood and a half-height composition with every face clearly visible. Keep facial features recognizable, match perspective and light direction, preserve natural body proportions, and use believable contact shadows. Show relaxed expressions with natural spacing between people. Do not add text, logos, extra people, or documentary claims. The finished image should be shared as AI-generated.
Ready to use. Add one clear portrait per person only in the image app you trust.
Copy this message into your usual group chat. It includes your chosen setup and the one thing each person needs to send.
Nothing is sent automatically. Share only with people you know.
Watch the Android app: choose a scene, read the photo tips, and prepare separate portraits. For a group photo, add one clear portrait of each person.
Silent recording of Android version 1.6.1, with screens held for readability. It shows preparation only, with illustrative AI portraits and scene thumbnails. No photo is generated in this clip. Screens may differ on iPhone.
Prefer scenes instead of prompt-writing?
Add one portrait per person and pick a ready-made vibe in TogetherLens. Photo generation uses paid tokens. Premium is optional. Check purchase options and prices in the app. This prompt is not imported into the app.Scan with your phone camera to open this scene and prompt. Your choices stay selected, and the app-store links are on the same page.
Open this setupThe code contains only your scene, mood, and framing choices. No photos or names.
WHY THESE DETAILS MATTER
Say that each portrait is an identity reference. Without that, a generator tends to invent extra people or blend two faces into one.
Portraits taken in different rooms carry different light. One stated direction and color makes the combined scene believable.
Three to six people keeps faces readable. A bigger gathering is usually better as two pictures than one crowded frame.


BEFORE YOU GENERATE
QUESTIONS
Describe every portrait as an identity reference, choose one specific place, ask the generator to match one light direction and one camera height, keep each face visible, and exclude extra people, text, and logos.
No. It does not ask for names, an account, or any photo. It builds text from your scene, mood, and framing choices in the browser.
Three to six reads best. Beyond that faces get smaller and identity detail gets less reliable, so a larger gathering usually works better split into two or three pictures.
No. Identity consistency varies by product and by how similar the source portraits are. Inspect the result carefully and avoid claims that any generator guarantees an unchanged face.