This is a first-and-last-frame video generation application based on LTX 2.3 Multi-Image Keyframe Guidance + PromptRelay Segmented Plot Control + Three-Stage Rendering Polish. Users only need to upload a 2x2 grid reference image, and the system automatically splits the image into four keyframes: top-left, top-right, bottom-left, and bottom-right. It then uses LTXVAddGuideMulti to inject these keyframes into different timestamps of video generation, achieving continuous video generation from the first frame to the middle transition and finally to the last frame. Compared to ordinary single-image-to-video, its core advantage is: instead of relying on just one image for the model to improvise, it uses multiple keyframes to control plot progression, character positions, and visual changes.
This workflow also combines three-stage sampling, Latent upscale, tiled decoding, and PromptRelay long prompts, making transitions between multiple images more stable. It is suitable for character transformations, continuous plot shots, first-and-last-frame transitions, four-stage story videos, and AI short drama storyboarding. When using it, users are advised to focus on controlling three things: the grid reference image determines visual changes, the segmented prompts determine plot actions, and the keyframe guidance strength determines transition stability.
🎁Claim RH Coins first, then experience the workflow! Click your profile picture in the top right corner → Invitation Code → Enter [rh-v1111] to instantly get 1000RH Coins, and log in daily to get another 100 RH Coins~ For more ComfyUI workflows, tutorials, and gameplay, follow our official WeChat account (AIKSK). Synchronously updated on Douyin / Bilibili / Xiaohongshu / YouTube (AI-KSK) ✨


No creations yet

No creations available.