

Standing portrait with a slight turn
A standing person turned a quarter step so the shoulders and torso read a new angle. The face, hair, and outfit stay where the camera put them, and the room behind the person stays readable.
Upload one photo, pick standing, sitting, or a slight turn, keep face and clothes, then compare and download.
Upload one photo, pick standing, sitting, or a slight turn, then inspect the split view before downloading.
JPG, PNG, or WebP · up to 5 MB · 256–4,096 px per side · ≤ 16 MP
Face, clothes, subject count, and the original scene are asked to stay the same on every run. This is not a skeleton editor and not a copy of a second pose photo.
Your pose-changed photo will appear here
Upload a supported photo. After it lands, the workbench checks its size and prepares one pose change.
Pose cases
These illustrative, AI-generated examples show a standing portrait given a slight turn, a full-body figure redrawn into a seated pose, and the same outfit held through a stance change. They are not measured outputs or a performance benchmark for the production service. Actual results vary, so inspect your own result against its source before you keep it.


A standing person turned a quarter step so the shoulders and torso read a new angle. The face, hair, and outfit stay where the camera put them, and the room behind the person stays readable.


A full-length photo restaged into a natural seated pose on a simple chair that fits the scene. Clothes, face, and subject count stay in place so you can judge the new stance without a second shoot.


A person in one set of clothes moved from a casual stand into a more open stance. Jacket edges, trousers, and shoes are asked to follow the body so the outfit still matches the original photo.
What changes and what stays
An AI pose changer is a photo editor that redraws stance on an existing person photo. This page changes pose. It does not promise an exact joint match, a copied reference pose, or a new character.
Every run carries the same locked instruction: keep identity, face structure, hair, clothing, accessories, subject count, crop, camera, and the original environment. The brief also says not to invent extra people or a new room.
Standing uses a natural weight-on-feet stance. Sitting uses a seated pose and asks for a full-body or three-quarter photo. Slight turn rotates the torso about a quarter turn. There is no free-text pose field and no second pose-reference upload in this version.
This is not a skeleton editor and not a guarantee that fingers, hems, or shoes will match the source joint for joint. Hands, limb count, and garment structure can drift. The output is a new raster image, so inspect it against the source before you use it.
Practical workflow
Start with the clearest photo you have. One person, in good light, with the face and as much of the body as the preset needs, gives the identity lock and the pose change the most to work with.
Choose a JPG, PNG, or WebP file up to 5 MB, with each side from 256 to 4,096 pixels and no more than 16 megapixels. After upload, the workbench checks those dimensions locally and rejects out-of-range photos before any credits are spent.
Standing is the default. Sitting needs a full-body or three-quarter photo and says so before you spend credits. Slight turn is the smallest change. This page does not accept a second pose-reference photo or a skeleton overlay.
Sign in, confirm the fixed cost of 20 credits, and start one tracked job. The result is a 1K image at the same aspect ratio as your photo. Controls lock while the job runs so the request and the result always describe the same pose.
Drag the split view across the face, hands, limb count, clothes, and any text in the frame. Download only when the drawing still agrees with the person in the photo. Retry the original task from the workbench if a connection drop pauses status checks.
The pose changer generates a new photograph from your upload plus a server-owned pose brief. It does not copy a stick figure, match a second pose-reference photo, or guarantee that a seated pose will invent missing legs from a tight headshot. Small details can move. Fingers, garment edges, and shoe placement are the usual places to look first.
Keep the original file. If a face, tattoo, logo, or piece of jewelry drifted, discard the result and try a clearer photo or a different preset. Multi-person photos keep each visible person in place, but likeness and pose for every figure is not guaranteed. If you need an exact copied pose, photograph the person in that stance.
Upload only photos you have the right and consent to process. Signed-in pose-change tasks and results can appear in My Creations so a task can be recovered after a temporary connection problem instead of being paid for twice.
The current storage backend has no automatic expiry for uploaded or generated image objects. Delete in My Creations calls the authenticated creation-delete endpoint and soft-deletes the task record from account history; it does not delete the underlying image object or provider copy. Review the Privacy Policy and contact support before uploading photos of children or other sensitive portraits.
Choose the right pose workflow
The useful split is what you supply and what you need back. This page takes a person photo plus a standing, sitting, or slight-turn preset. A new-person pose generator invents someone in a pose. A skeleton pose editor copies joints from a stick figure or a second pose photo, which this page does not accept.
Tool specifications
The workbench accepts one JPG, PNG, or WebP image up to 5 MB. Each side must be between 256 and 4,096 pixels, and the image must stay within 16 megapixels. One run costs 20 credits and returns a 1K result at an automatic aspect ratio. Higher output sizes are not offered on this page because the automatic aspect ratio is only available at 1K.
Processing infers a body stance that the camera never recorded, so the result is an interpretation rather than a studio recapture. By using the tool, you confirm that your upload complies with the Terms of Service. Reviewed by the Banana Pro AI product team · Last reviewed: 2026-09-12.
Keep editing
Use the narrowest tool that fits. The image editor for a broader instructed change, a headshot when you need a professional portrait, or a character generator when you are inventing a figure rather than posing a photo you already have.
Questions
Clear answers about what this tool is, what stays in the photo, skeleton editors, sitting shots, credits, and privacy.
An AI pose changer is a photo editor that redraws the body stance of a person already in a photograph while asking the model to keep the same face, clothes, and number of people. This page is that editor, with standing, sitting, and slight-turn presets on one uploaded photo.
That is the goal of the locked brief, which asks the model to keep face structure, hair, clothing, accessories, subject count, crop, and the original scene. It is not a guarantee. Clear, well-lit photos hold identity best. Compare the result with the split view before you use it.
No. This version is text-guided. It does not accept a pose-reference image, a stick-figure overlay, or a free-text pose field. If you need an exact copied pose, photograph the person in that stance or use a skeleton pose editor.
Sitting needs a full-body or three-quarter photo so the seat, hips, and legs have something to draw from. The workbench says so before you spend credits. A tight headshot does not give the model enough body to seat without inventing limbs, and that invention is not a promise.
One run uses 20 credits and returns a single 1K image. Files must stay within 5 MB and 16 megapixels. Signed-in tasks can appear in My Creations. There is no automatic expiry. Delete in My Creations soft-deletes the task record; it does not delete the underlying image object. Review the Privacy Policy before you upload.
Keep the original. Check the hands.
Start with one clear photo, review the fixed 20-credit cost, and compare the pose-changed result with the original before you keep it.
Open the pose changer workbench