Your photo — or theirs, with permission. Same face, new pose — and somehow the whole temperature of the room changes. Shoulders settling into stillness. The looseness right before someone leans back. A stride caught in low light. Lamplight through a gap in the curtains, the hush after midnight, a look that finds you across the room — and your pulse quietly agreeing. 55 poses, 4 camera angles, 3 framings. Tonight, let that photo hold a shape you've never seen it hold.
JPG, PNG, or WebP — full body in frame, soft, even light. This photo sets the person and the mood. Use one you took yourself, or one you're allowed to edit.
Choose one of 55 presets, then set the camera angle and the framing (full body / mid-thigh / waist-up). The prompt field is optional — it refines details.
Check the credit estimate, then generate. Before and after, side by side — face, hands, the line of the body, the background — at full size before you save. When the light lands, your pulse will know before you do.
AI-generated illustration
Every creation combines a pose preset, a camera angle and a framing. The framing also sets the output size. The workspace also includes a clothing-state control, but pose is always the star of this page.
Three renders out of one poolside selfie: the same fictional adult woman — shoulders at rest, fingertips loose on the armrest, one stride catching the low light. Same person, three different temperatures. Drag the slider and see for yourself.
Pose reference
Keep the whole person in frame when choosing a full-body pose. A clear view of the arms, legs and outfit gives you a useful starting photo.
Pose reference
Browse the visual pose cards instead of describing every joint. Choose a reference, then review its preview before using your credits.
Pose reference
Use the optional Prompt field for an expression, outfit detail or setting. Keep your note focused so the photo and selected pose remain the main references.
All three sets are original AI illustrations featuring three different fictional adult women. Within each set, the source photo and the result are the same person. They exist to explain the flow — they are not outputs of a real generation task and guarantee no outcome.
This is an AI pose changer that re-renders a photo you already have. Instead of drawing from a blank prompt, your photo decides the person, the light and the background, while the pose preset, camera angle and framing decide the new composition. A clothing-state control also lives in the workspace, but pose is always the star of this page.
Use a full-body photo with even light and no heavy shadows across the body or legs. Front or 45-degree angles read best — you can see both the person's outline and the background. Upload the original JPG, PNG or WebP file, not a screenshot or a messaging-app copy. Poses with overlapping hands and feet, or objects in front of the body, make the new composition harder to read.
The presets split into four categories: 16 standing, 3 sitting, 5 movement and 31 floor poses. Standing and sitting keep the body upright and read most reliably; movement adds energy but wants a wider framing; floor poses drop the body low and change the background more. There is exactly one upload slot and no way to import an external pose image — every pose comes from this page's presets.
Beyond the pose there are two more controls: camera angle (eye-level, low-angle, extreme-low, overhead) and framing (full body, mid-thigh, waist-up). Framing also sets the output size: 1024×1368 for full body, 1024×1280 for mid-thigh, 1024×1024 for waist-up. Eye-level is the most neutral, low-angle lengthens the body, extreme-low is the most dramatic, and overhead flattens the scene.
Uploading a photo and browsing presets costs nothing and starts no task. Creating needs an account and spends credits, with the estimate shown before you submit. There is no "completely free" or "unlimited" tier on this page. A retry after a failed attempt is billed as a new creation.
This is an image-to-image editor: the person, light and background come from your file. Text alone can't start a task. Perfect replication isn't guaranteed — hands, accessories and body edges can shift slightly. The inspection order never changes: face, hands, body edges, background.
Pose, camera and framing carry the main instruction, so the prompt only refines: hand placement, an angle you want, a small gesture. One sentence, one idea. Change a single variable per attempt — pose card first, then camera, then prompt. Every attempt spends credits, so one variable at a time is the cheapest path to the result you want.
It's for people who want to see their own photo in another pose: building a board of composition ideas, previewing a shot before a shoot, or arriving from a search like "ai pose" or "change pose in photo". It doesn't cover poses outside the presets or imported pose references. Only adult photos you shot or have the right to edit — never photos of anyone who hasn't agreed. Commercial and public use calls are yours to make and own.
It changes the pose in a photo you already have. Upload one adult photo, pick a pose from 55 presets and generate — the same person is re-rendered in a new pose, with the face, light and background of your photo. You can also pick the camera angle and framing: full body renders at 1024×1368, mid-thigh at 1024×1280, waist-up at 1024×1024.
Adult photos only — ones you shot yourself or clearly have the right to edit. Photos of people who haven't agreed are not allowed. A full-body, evenly lit JPG, PNG or WebP gives the most stable result.
From 55 presets: 16 standing, 3 sitting, 5 movement, 31 floor. There's no upload slot for external pose images — people searching "change pose in photo" land on this preset-based flow. The prompt field only refines details; the preset always decides the pose itself.
Framing decides it: 1024×1368 for full body, 1024×1280 for mid-thigh, 1024×1024 for waist-up. The camera angle changes the viewpoint, not the canvas. The credit estimate follows the framing you pick.
Uploading photos and browsing presets is free and starts no task. Creating requires an account and credits, with the estimate shown before you send. There is no "completely free" or "unlimited" option.
It depends on the rights to your photo, the consent of the people in it, and the site's terms. Confirm the rights you need before commercial use; how you use what you create is your own responsibility. Read the latest terms as well.
Your photo supplies the look; the pose reference image marks the goal. Knowing the two roles makes picking photos and judging results much easier.
“you can also pass an initial image as a starting point”
Hugging Face's image-to-image guide describes the initial image as a starting point. For full-body poses, a photo with the whole body in frame makes the new composition far easier to compare.
The 2023 ControlNet paper studied human pose as a way to guide image generation visually. A reference for understanding pose conditioning. When editing, compare the finished image against the chosen reference — check face, hands and background before saving.
Zhang, Rao & Agrawala — Adding Conditional Control to Text-to-Image Diffusion Models (2023)
Page details reviewed:
Your photo — or theirs, with permission. Choose the pose; the rest is one button. Somewhere past the lamplight, that person is already holding a shape you've never seen — all that's left is to look.
Selected pose
Photo
Pose reference
Hand on hipFinished example
One photo, a brand-new pose. Drag the divider to compare.
AI-generated illustration
Photo
Pose reference
Hand on hipFinished example
One photo, a brand-new pose. Drag the divider to compare.
AI-generated illustration
