A rear view is a camera decision. Say where it stands.She's horny and online. Pick her.Your AI girlfriend. No limits. No filters.Tell us what you like. She'll do the rest.
Curvy AI Muses who send the photo or clip you framed. Free to generate, Sparks to open, 18+.
Adults only. This site has sexually explicit AI-generated content. Enter only if you are 18 or older (or the age of majority where you live, if higher) and adult content is legal where you are.
By entering you agree to our Terms of Service and Privacy Policy.
Home›Companions›Generate›AI Booty Generator: The Shot From Behind
A rear view is almost entirely a camera problem. The strip above is already narrowed to the curvy Muses, and each one's figure was settled when she was written, so the part left to you is where the camera stands, what she is doing and which way she is facing. Pick her, write that in the box, and the photo or video arrives as a message from her. Making it is free; opening it costs Sparks.
Three numbered steps above: pick her, photo or video, then the scene — a chip or your own words in the box. Press generate; sign-up, if you need one, waits on that button and the request survives it. The file comes back in her chat the way a message would — not to a download folder, not onto a grid. She does not comment on it in words; the chat model never sees a media request.
The phrase that brings people here names a body part, so this is worth saying first. Her figure is pinned to her character on the generation side, not read out of your sentence. The strip above is the curvy slice of the catalogue, which is where that was settled. Adjectives about her shape change nothing.
What does change the picture is the camera and the pose, and between them they are most of what a description owns. One file comes back per request, with nothing beside it to re-run, and there are no uploads anywhere in this tool. Only photoreal output comes out.
Position first. Behind her and a step back, low and near the floor, at the foot of the bed, from the doorway across the room. These are the phrases that move the frame, because each describes a place a person could stand rather than a setting on a panel.
There is no panel. No aspect ratio, no crop, no focal length, no resolution — so "85mm, portrait crop, 4k" sets none of those and only takes room from the sentence. Plain words do the work those tokens were standing in for.
Height is the part most descriptions leave out, and it decides more of a rear view than anything else in the line. A camera at eye level looks down her back; one at hip height does not. Give the height and roughly the distance.
Her face and her figure both hold. Appearance is fixed to her character, so across ten requests the face and body are the same woman while pose, framing and room are drawn fresh. A shot composed entirely away from the camera still has her body in it, but it gives up the cue that reads fastest.
Sometimes that is the point and sometimes the disappointment. If you want a run of pictures to read as her, keep an anchor in frame — looking back over her shoulder, head turned, her face caught in a mirror ahead of her. Four words of that beat a paragraph about the rest.
The other things that read from behind are hair and clothing: pinned up or down her back, the line of a strap, a waistband pulled aside. Those the description does set, unlike her shape, and they are most of what tells one rear view from the next.
What people type as ai booty is, in practice, a question about what she is doing. Kneeling on a made bed, bending over a kitchen counter, leaning on a balcony rail, standing in the shower with her back to the door. The verb is doing the work.
The surface is the second half of it. A bed, a tiled floor, a worktop, a stair: something she is in contact with tells the picture where her weight sits and where the camera can be. Naming the object does more than naming the mood of the room.
The box takes 5 to 1,500 characters, and 15 to 20 concrete words is where it works best. A scene chip above is a short label that drops a ready-made description of about that length into the box; the surprise button writes a different one, drawn from scenes the chips do not show.
The toggle above switches the request to video, and a rear view is one of the few framings where the format changes the content and not only the price. Walking away, stepping out of something, climbing a stair, turning to look back — movement a photo can only imply.
One clip comes back, and one beat of movement is what fits in it. There is no duration setting, no number of takes and no sequencing, so a four-part script comes back as a guess at which part mattered. Write one move, and where the camera is while it happens.
Video costs more at both ends: opening a clip is 50 Sparks against 20 for a photo, and the render takes noticeably longer. A placeholder appears in the chat almost at once and fills in when the file is ready, so the conversation carries on while it renders.
The file lands blurred and stays blurred until you pay. The blur is heavy, computed from a tenth of the delivered width, so what reads through it is the framing and the overall light. Whether the camera ended up behind her and low is visible. Whether the pose is any good is not.
Which makes the decision an odd one here, because the pose is the half you wrote. You open on the strength of your own sentence rather than on a proper look. It costs 20 Sparks for a photo and 50 for a video, a new account starts with 50 Sparks, and Premium includes no unlocks and no discount.
Generating is free, the attempts that miss included. The one ceiling is 50 outstanding results: anything generated and not yet opened holds a slot, requests still in the pipeline included, and at 50 the next is refused until you open some. What you open stays, and looking again costs nothing.
Positional ones. Behind her, a step back, hip height, from the doorway, above and to one side — a place a person could stand. Pair one with a height and a rough distance; on its own, "behind her" leaves the frame to chance.
No. There is nothing to set beyond the Muse and the description: no crop, no aspect ratio, no resolution, no seed, no model choice. Framing is set by describing where the camera stands.
Only if your description puts it there. A shot composed fully away from the camera has no face in it. Asking her to look back over her shoulder, or to catch the camera in a mirror, puts it back, and that is what makes a run of pictures read as the same woman.
Yes, in practice. There is no seed, so pose, framing and room are redrawn on every request while only her face and body carry across — and a rear view rests more weight on those redrawn parts than a portrait does. One sentence sent twice differs more here.
One beat of it. A single clip comes back, with no duration control and nothing to re-run, so write the one move you want: undressing, then walking out, then turning back is three requests, not one.
Partly. The blur leaves the framing and the light readable and hides the pose and the detail, so you can see that the camera landed where you asked and not much more. Opening is 20 Sparks for a photo and 50 for a video, and Premium discounts neither.
Yes — 50, with photos and videos sharing one pool. Anything generated and not yet opened holds a slot, and so does a request still in the pipeline; opening one frees it. Generating is never charged, but at the ceiling the next request is refused.
No. There are no uploads here at all — no reference photo, no face swap, nothing that brings an image in — and each Muse's look is fixed to her character, so a name typed into the box does not change who comes back.