Overview
This workflow uses Wan2.2 I2V together with SVI 2.0 PRO LoRA to generate long-form videos of 81 frames or more.
Instead of generating 81+ frames in a single pass, the video is created in multiple segments and then merged.
Compared to traditional extend/long workflows, this approach preserves motion information from the previous segment, resulting in much smoother transitions without noticeable seams.
Core Features
Lightx2v-only Mode
Offers high prompt adherence and fast generation speed, but may introduce the characteristic slow-motion effect typical of Lightx2v.
3-pass Mode
Slower generation speed, but capable of producing more dynamic and realistic motion in real time.
Optional Features
Resume Mode
Allows you to continue generation not only from the first frame, but also from a previously generated video.
If something goes wrong partway through the process, you can save the generated segment and resume from that point without starting over.
Recommended Settings (Updated: 2026-01-10)
Use 3-pass mode with the following settings:
Normal step count: 1
Lightx2v step count: 3
CFG: 3.5 - 5.5
Lightx2v LoRA:
lightx2v_I2V_14B_480p_cfg_step_distill_rank256_bf16
High: 3.0 / Low: 2.0Pusa: Enabled
I tested many different configurations, and this setup produced the most dynamic motion while maintaining natural motion speed and avoiding artifacts.
If the output starts to look like slow motion, try increasing the CFG value by 1.0 at a time.
Notes
This workflow is configured for the GGUF version, but fp8 and Distill versions can also be used by switching the model loader.
Designed with 480p output in mind.
There are many variations of Lightx2v LoRA, and their suitability depends on the type of video you want to generate.
If visual artifacts appear in 3-pass mode, try lowering the CFG value from 8.0 to around 6.0.
Acknowledgements
Special thanks to Kijai for providing the excellent LoRA models and custom nodes.