Both source pictures and the composite below are genuine output from this app's own two runs. Nothing here is a mock-up and nothing here costs you anything.
The brief this was painted from
Before you spend anything
Open any of these. Nothing in here runs anything or costs anything.
How does it blend two photos into one?
In two steps, and the app shows you both. First a vision run looks at your two photographs and writes a render brief describing the finished picture — what is kept from one, what is borrowed from the other. Then an image model paints that brief. The brief is visible and editable in between, so the expensive half is only ever spent on wording you have read.
Does the image model see my actual photographs?
No, and this app says so rather than implying otherwise. On this platform the image models reject file attachments outright, so the renderer works from the written brief alone. That is a real limitation — the composite is a new picture in the spirit of your two, not a pixel-level montage of them. It is also the reason this app cannot face-swap: there is no path by which a photograph of a person reaches the renderer.
The two photographs are input to the reading run through the platform's vision path, not retyped as a text description. Only the brief crosses to the renderer.
What will it refuse to blend?
Four things, refused in your browser before any credit is reserved.
- Identity transplant. Carrying one person's face or likeness onto another person or another person's body — face swaps, however they are worded.
- Fabricated records. Asserting who someone is and placing them in a place, an act or a company they were not in.
- Sexualisation. Sexual, intimate or undressed pictures, including making an uploaded subject more sexual than the photograph they arrived in.
- Minors. Anything suggestive, romantic or aged-up around a subject who could read as a child.
Merged-subject and double-exposure modes are also refused outright when the app sees a person in either photograph, because those two modes become a face swap the moment a person is in frame.
What happens to my photographs?
They are resized to 1280 pixels on the long edge inside your browser before anything is uploaded — typically a fifteen to thirty times reduction. The re-encode also discards the EXIF block, so the GPS coordinates baked into a phone photograph never leave your machine. The resized copies go to your own app storage on your account and are shown back through short-lived signed URLs.
What does one blend cost?
Two metered runs, and the app quotes both reserves before either one. The reading run reserves about sixteen cents and the rendering run about twenty-seven — figures read live from the platform's own estimate rather than repeated from documentation. Both are reserves, not prices: a settled charge on this platform varies several-fold against an identical reserve, which is why the app quotes the reserve and never a fixed price. You only ever reach the second reserve after reading the brief the first one produced.
Can I use it without signing in?
Every part that does not spend credits works signed out: both upload slots, all six blend modes, the strength dial, the consent guard, the live price meter and the bundled example with its real brief and real result. Guest sessions carry no balance, so the app says so at the run button rather than failing after you press it.
Where does a finished blend go?
Onto your screen first, with a PNG download, a copy-the-brief button and an open-full-size button beside it. The composite is then stored on your own account and shown through a short-lived signed URL, so it reopens from History on any device you sign in on. Past blends are searchable by meaning rather than by keyword — ask History for “the warm lake one” and it finds it.