What’s new

Click a reference image to see it full size, in the prompt box (Image or Video) and in Edit AI Image. With several references, use ← → or the side buttons to flip through them; Esc or a click anywhere closes it. Try it →

The Library adds 33 poses: 12 women's model and dance poses and 21 men's full-body poses, each from five angles as line art and depth maps. They come first in the Pose tab, which now has 321 poses, and there are new Walking & running and Gestures categories. Try it →

You can now email us: there's a contact link in the footer.

The Library adds Muse originals: 181 faces of women and men from around the world and 100 men's outfits, shown first with a Muse badge. The Face and Outfit tabs have a new gender filter, and Face now shows close-ups only. Try it →

New Guides page: browse every guide, newest first, by category or tag, or search them. Guides not yet translated appear in English with a language badge. Open it from Guides in the top bar (a book icon on phones) or the footer. Try it →

The 🎲 dice now also picks faces from the Muse originals. On the English site it picks only from those; in Japanese and Chinese it picks from all women's faces in the Library. Try it →

Image to Video: pick one photo and bring it to life as a video, with a preview right on the page when it's done. Open it from the footer or from the top of Create AI Video. Try it →

New guide on generating AI images: writing prompts, choosing style and aspect ratio, using face, outfit and pose references, editing, and fixing common problems. Example prompts have a Copy button. Find it in the footer. Try it →

Broken or mistyped links now open a proper “page not found” page in your language, with links back to the homepage and the tools, instead of an error.

New Updates page: see each release's new features, improvements and fixes, and jump straight to the page where you can try each one. Find it in the footer.

The History tab in Edit AI Image now shows only the image you're editing and its versions: the upload or creation it started from and every edit made from it. Try it →

When you come back to Edit AI Image, old versions whose images have expired no longer show up. Try it →

Drop an image file on a reference to replace it in place. “Image 2” keeps its number, so you can swap images even when all 3 slots are full — in the prompt box and in Edit AI Image. Try it →

Edit AI Image keeps the shape of Image 1 — edits no longer come back cropped to 16:9. Try it →

360 Orbit: turn one square image into a 3-second video where time freezes and the camera circles your subject. Non-square images get a crop frame. $0.10 per second. Try it →

Create AI Video: turn a prompt, or one reference image, into a 1–15 second video. $0.10 per second, charged only when the video is delivered. The homepage prompt box has a new Image / Video switch. Try it →

Library: 1,499 faces, 2,727 outfits (photo and anime) and 288 poses in five angles. Search in English, Japanese or Chinese, ★ your favorites and send up to 3 to the prompt box. Try it →

Hide images in My Creations: a hidden card stays covered — no image, no prompt — until you click it. Try it →

Generations in progress show an estimated time with a live countdown.

My images is now My Creations, with separate Images and Videos tabs. Try it →

The 🎲 dice uses depth-map poses, so the pose comes through more faithfully. Try it →

Edits and dice rolls keep the aspect ratio you pick instead of the model's default. Try it →

Muse launches: edit any photo or create images from text with an uncensored model — 1,300+ styles, the 🎲 random face, outfit and pose, and 3 free generations for new accounts. Try it →