Turn one source image or two keyframes into a directed video sequence. Describe the motion, choose the format and frame count, then generate without building a manual animation timeline.
Source image *
JPG · PNG · WebP · 10 MBJXP 2.0 sample output
Upload a frame, describe the motion, and generate your own result.
JXP 2.0 is an AI video model that turns image references and written instructions into moving visual sequences — in single-image mode or with start-and-end keyframes.
Instead of placing animation points on a timeline, you describe subject movement, camera behavior, pacing, and lighting in a prompt — then JXP 2.0 builds the sequence from your uploaded frame.
Use it to create a shot, motion study, transition, or visual draft from prepared images, then review, revise, and bring the result into your production workflow.
Browse sample outputs across cinematic motion, storytelling, and social-ready formats. Hover a tile to preview the clip.
Each control corresponds to a practical decision: where the shot begins, where it ends, how it moves, how long it runs, and where the final format will be used.
Begin with a product photo, character portrait, illustration, or composed scene. JXP 2.0 uses that frame as the visual anchor while your prompt defines what moves, where the camera travels, and how quickly the action develops.
Upload a start frame and an end frame when the destination matters as much as the opening. Describe the relationship between both images, then let the model generate the transition.
Write the movement you want to see instead of building an animation timeline. Name the subject action, camera direction, speed, lighting, and mood in one focused prompt.
Create landscape, vertical, square, or portrait video without cropping a finished result into the wrong composition. Plan for a website hero, social feed, or product page from the start.
Select 81, 121, 161, 241, or 441 frames. Every option follows the model requirement that total frames use an 8n + 1 value. With playback fixed at 24 fps, a higher frame count gives the action more time to develop.
Leave the seed empty for a new starting point or enter a positive integer when you want a more controlled comparison between prompt variations.
Single Image mode is the direct route when you want to preserve one composition and add movement. Keyframes are better when a transition needs a defined destination.
Use a photo or illustration as the opening visual. Describe subject motion and camera movement while the source image anchors the scene.
Set both ends of the shot. Your prompt explains how the model should move from the first supplied frame to the second.
Three steps from a still frame to a review-ready clip — all inside one creation area.
Choose Single Image or Keyframes, then upload one source image or a matched start-and-end pair. Use clear images with the subject at a useful scale and enough room for the movement you plan to describe.
Write the main action, camera direction, pacing, and lighting in the prompt. Pick the aspect ratio, frame count, and an optional seed before you render.
Submit the task and follow progress in the preview panel. Download the result or start another pass with a tighter prompt, different frame count, or revised seed.
Use it when a still visual already communicates the design but the next deliverable needs movement, pacing, or a transition.
Turn a still product image into a short display clip. Describe a slow rotation, a controlled camera push, moving reflections, or a reveal between clean start and end frames.
Build vertical 9:16 ideas for Reels, Shorts, and mobile campaigns. Test framing and movement before committing time to a longer editing workflow.
Add a restrained gesture, a shift in expression, environmental motion, or a camera move while keeping the supplied artwork as the visual foundation.
Connect storyboard panels with keyframe mode to explore how one shot could lead into the next. The result can help communicate pacing and camera intent.
Animate a prepared app screen, device composition, or branded graphic for a landing-page loop, feature reveal, or presentation opener.
Use clearly different start and end frames to plan transformations, time changes, location shifts, material changes, and other transitions that need a defined destination.
JXP 2.0 is an AI video model for turning a source image and a written motion prompt into a video. It supports both single-image animation and keyframe workflows.
Upload one image, describe the movement and camera direction, choose an aspect ratio and frame count, then start the task. JXP 2.0 uses the image as the visual starting point for the generated sequence.
Keyframe mode accepts a start image and an end image. Your prompt describes the transition between them, while the model builds the motion that connects both frames.
The generator offers 16:9, 9:16, 1:1, 4:3, and 3:4. These cover landscape video, vertical social clips, square posts, and traditional portrait or presentation formats.
This page uses a fixed 24 fps frame rate. A higher frame count produces a longer sequence, while lower frame counts are useful for shorter motion tests.
A seed gives the generation process a repeatable starting value. Reusing it can help when testing prompt changes, though generated motion can still vary when other inputs change.
Describe the subject action, camera movement, pacing, lighting, and the visual relationship between the first and final frame. Use direct instructions and keep competing actions to a minimum.
Yes. When the task finishes, the result appears in the preview panel with a download action so you can save the generated video.
Upload the frame that already captures your idea. Add the motion, format, and timing you need, then turn it into a video you can review and refine.
Open the JXP 2.0 generator