2026 0419: Added a node for restoring colors, because in rare cases, the previous model caused color changes in videos during generation. Added it for now. Enable it if needed.

2026 04 17: Initially planned to update and optimize, but found that the smooth models v20 and v30 were not suitable for creating the first and last frames. The quality of the first and last frames in other models was not as stable as smooth's. So, I only made slight adjustments to the workflow without any major breakthroughs. Tested it all afternoon and felt that unless there's new material, there's nothing much to improve. If you need the previous AI applications, you can use this workflow of mine. Just modify the save nodes and your desired prompts.

LU_Short Series Six-Pack Part 4: First and Last Frame Video Generation
High-dynamic generation of first and last frame videos. You can let AI assist in generating prompts or use your own prompts.
Default: AI-generated prompts.
If you input your own prompts, just define the prompts and select false:div prompt.

ps: I generated videos at 1280*720 resolution, with 4 seconds of video taking about 4 minutes (03:48/03:19).
ps: The default fps is 16. To modify it, use the workflow, or I might later release a small application for frame interpolation and upscaling.

LU_Short Series Six-Pack Part 1: Text-to-Image
https://www.runninghub.cn/post/1987052252043087873
LU_Short Series Six-Pack Part 2: Image-to-Storyboard
https://www.runninghub.cn/post/1988519232235331585
LU_Short Series Six-Pack Part 3: Image-to-Video
https://www.runninghub.cn/post/1987043435041460226
LU_Short Series Six-Pack Part 4: First and Last Frame Video Generation
https://www.runninghub.cn/post/1987043353088954370
LU_Short Series Six-Pack Part 5: Multi-Frame Video Generation
https://www.runninghub.cn/post/1987043344478089218
LU_Short Series Six-Pack Part 6: Video Synthesis & Audio Synthesis.
https://www.runninghub.cn/post/1987052320276066306