This is a workflow for generating long-form videos using the Wan2.2-Remix model. I've created two versions—T2V and S2V—to meet different user needs. A big thank you to FX_FeiHou for creating the Wan2.2-Remix_I2V_V2.1 model. You can download the model here: https://civarchive.com/models/2003153/wan22-remix-t2vandi2v. For important details, please see the workflow usage instructions below.

💻I've already set up both the ➡️ TI2V and ➡️ S2V workflows for you. Just click the links to use them online.

🎁Bonus: If you're signing up for the first time, you can get 1,000 free RH Coins by using my link and the invite code ➡️rh-v1182. Plus, you'll get another 100 RH Coins for logging in daily.

🚀Workflow Usage Instructions:

1. TI2V: This version uses auto-generated prompts combined with a "for loop" to generate a 30-second high-dynamic video. All you need to do is load a single image, and it will automatically output a video script and generate the video. The downside is that character movements and scene transitions rely entirely on the prompt description, so there's some uncertainty. Of course, you can also use custom prompts to describe the script for each shot to get closer to the specific effect you want.

2. S2V: This version uses storyboard images to generate a series of continuous video clips with smooth action, which are then merged into a single 30-second video. This means you need to prepare or generate storyboard images first, using them as the start and end frames for each segment. It uses both the storyboard images and prompts for dual control, allowing you to better tailor each segment to your needs. However, the overall coherence isn't quite as good as the T2V version, and you might see noticeable seams where two clips join.

3. In the S2V version, I've included a preset "Generate storyboard images" group to help you create the individual storyboard shots. In my testing, I found that using Gemini 3 Pro Preview to write the scripts and Nano banana Pro to generate the storyboard images gave better results, but feel free to experiment with different methods.

4. I've also included preset groups for video frame interpolation and video upscaling. These aren't required to run, so they are set to be ignored by default. You can enable and use them as needed after generating the full video.