Image to video
Your photo, four seconds later.
Upload an image, describe the motion, and the studio animates it - Seedance 2.5 for overall quality, Kling 3.0 Turbo Pro for hi-fi motion, Veo 3.1 for cinematic polish, or the unrestricted Wan variants when mainstream models refuse the source. First/last-frame control included for precise moves.
Free messages to start · no card · no identity attached
The image-to-video lineup
Pick the animator by what the shot needs.
Seedance 2.5
Best OverallThe default: pin a character, a wardrobe, a location and a style plate - up to 30 reference images per shot - before anything moves.
Kling 3.0 Turbo Pro
Hi-FiThe motion specialist: when the animation has to hold up - bodies, fabric, water - Kling keeps the physics honest.
Veo 3.1
4KCinematic polish for stills that need to become camera footage - reads technical direction best.
Wan 3.0
New · UnrestrictedA continuous take up to 30 seconds from one still, with no provider-side prompt moderation.
Wan Spicy line
18+ gatedThe unrestricted image-to-video variants for legal briefs mainstream tools refuse - behind their own 18+ gate.
Happy Horse 1.1
Most UnrestrictedThe least filtered animator in the Studio, with widescreen support.
The still-first workflow
The cheapest fix for bad video is a good image.
Composition, lighting and the face are decided in the still - the model only has to move them. The Studio is built around that order.
Get the still right
Generate the frame in the Image Studio or upload one. Every fix is cheaper here than in a re-rendered clip.
Image StudioPin the references
Seedance 2.5 accepts up to 30 reference images per shot, and Character Studio (beta) keeps the same face from clip to clip.
Character StudioAnimate it
Describe the motion, pick the model and duration. Set the first frame, optionally pin the last, and chain clips into longer moves.
Image to VideoFinish it
Cut clips together in the Movie Editor and upscale the final piece 2x or 4x for delivery - all in the browser.
Movie EditorFree credits on sign-up · price shown before every render
Why image-to-video gets refused
The strictest moderation in AI is pointed at your uploads.
Image-to-video is where hosted tools moderate hardest - the input image and the output frames both get reviewed, and anything with skin, real people, or edgy aesthetics trips the classifier. Your own photos become the thing you're not allowed to animate.
“This image violates our policy”
Portraits, art, fashion shots - upload moderation bounces ordinary source material, and every tool draws the line somewhere different and invisible.
Your uploads, their archive
The images you feed a hosted animator sit on someone's servers, attached to your account, governed by terms you didn't read.
Briefs people bring to the studio
Here's how OpenGradient is different
Animate on your terms
Frame-level control
Set the first frame, optionally pin the last frame, and chain clips into longer moves - supported models are marked in the picker.
An unrestricted lane
Wan image-to-video variants (including the Spicy line, 18+ gated) run without provider-side prompt moderation - for the legal briefs mainstream tools refuse.
Private uploads
The anonymity layer separates you from your generations, and share links expire after 24 hours.
How privacy worksFree messages to start · no card · no identity attached
The anonymity layer
Animate the photo without donating it.
Not a policy promise - a protocol. No single party, including us, ever holds both your identity and your question.
Read the full security architectureEncrypted on your device
Messages are encrypted locally before they leave your browser.
Anonymized by Oblivious HTTP
Network relays split who you are from what you're asking. No one sees both.
Unlocked only in a secure enclave
The model reads your prompt inside attested hardware that never learns who sent it.
Questions, answered
Frequently asked
Which model should I pick?
Seedance 2.5 is the best-overall default. Kling 3.0 Turbo Pro excels at high-fidelity motion, Veo 3.1 at cinematic polish. The Wan variants (including the 18+ Spicy line) and Happy Horse 1.1 are the unrestricted lane. Prices per clip are shown on each - auditioning two models on the same image costs cents.
What resolution and length can I get?
Model-dependent - up to 1080p output, with clips from 5 up to 30 seconds on Seedance 2.5 and Wan 3.0, extendable further by chaining with last-frame control. Upscaling to 2x/4x is built into the studio for delivery.
What are the rules for people in photos?
Animating yourself, consenting subjects, or fictional/artistic characters is fine. Sexual content depicting real people without their consent is prohibited, full stop - the 18+ gate does not change that.
Pick a photo. Give it a verb.
Free to start, no card. Upload an image and see it move in under a minute.
Unrestricted means no provider-side prompt moderation on the open models - not no rules. Illegal content is banned, explicit capabilities are gated 18+, and sexual content depicting real people without their consent is prohibited.