<div data-youtube-video="https://youtu.be/R3W2ab1_TzY"><iframe src="https://www.youtube.com/embed/R3W2ab1_TzY" width="853" height="480" allowfullscreen></iframe></div> 
 
RunningHub workflow page: https://www.runninghub.ai/zh-cn/post/2098788728777388033?inviteCode=rh-v1111Bilibili tutorial video: https://www.bilibili.com/video/BV1vFYz6HECU/# SeedVR2 and source-FPS video post-processing## Purpose and preparationResolve the matching custom-node/runtime versions and assets. Load authorized media and inspect active inputs, actual schedules, effective dimensions and every save destination before a small test.This file processes loaded video; it is not an H3 generation model route. Loaded frames enter SeedVR2 with linked new_resolution 2160, batch 1 and a randomizing seed. No input-dimension multiplier or separate 3840×2160 normalization node exists. The 2160 control is not proof of a particular output raster.CreateVideo receives loaded source audio and VideoInfo source_fps output 0, not loaded_fps output 5; its old 30fps widget is overridden. Loader force_rate0,skip0,every_nth1,cap0. If you alter loading/resampling, review playback timing instead of assuming the source-fps link remains appropriate.SaveVideo 20 writes the pre-interpolation output and passes VIDEO to interpolation 17. SeedVR2 images also connect to its image input; input precedence requires matching implementation. Interpolator VIDEO output 1 reaches SaveVideo 22. Both saves are active. Keep-audio does not prove bit-perfect sound, VFR correctness or successful frame interpolation.Interpolator requests 2x,motion auto,threshold0.24,keep_audio on; runtime_dir/wine_prefix are empty. Dependencies, actual source-FPS multiplication and hardware support are untested.## Files and selected assets- [Workflow JSON]()- [Selected cover]()- `seedvr2_ema_3b_fp8_e4m3fn.safetensors`## Version, evidence and rightsThis is a local descriptive draft, not a published upgrade. Files01–08 are byte-identical to originals; only the standalone ninth file's historical video-preview field is cleared in its new copy. Originals remain unchanged.No model inference, visual/audio acceptance, wall-clock benchmark or license verification was performed. Covers are selected assets, not accepted render evidence. Weights and external input media are not bundled. Check workflow, nodes, models, LoRAs, source media and cover rights separately; no blanket commercial license is asserted.[Source-provided tutorial](https://www.youtube.com/watch?v=R3W2ab1_TzY) · [Source-provided workflow entry](https://www.runninghub.cn/post/2098776201618878466/?inviteCode=rh-v1111)Links are not live-verified. The workflow is supplied through AI-KSK.

Universal Video Upscaling SeedVR2 with DLSS Frame Interpolation

<div data-youtube-video="https://youtu.be/R3W2ab1_TzY"></div> 
 
RunningHub workflow page: https://www.runninghub.ai/zh-cn/post/2098788728777388033?inviteCode=rh-v1111Bilibili tutorial video: https://www.bilibili.com/video/BV1vFYz6HECU/# SeedVR2 and source-FPS video post-processing## Purpose and preparationResolve the matching custom-node/runtime versions and assets. Load authorized media and inspect active inputs, actual schedules, effective dimensions and every save destination before a small test.This file processes loaded video; it is not an H3 generation model route. Loaded frames enter SeedVR2 with linked new_resolution 2160, batch 1 and a randomizing seed. No input-dimension multiplier or separate 3840×2160 normalization node exists. The 2160 control is not proof of a particular output raster.CreateVideo receives loaded source audio and VideoInfo source_fps output 0, not loaded_fps output 5; its old 30fps widget is overridden. Loader force_rate0,skip0,every_nth1,cap0. If you alter loading/resampling, review playback timing instead of assuming the source-fps link remains appropriate.SaveVideo 20 writes the pre-interpolation output and passes VIDEO to interpolation 17. SeedVR2 images also connect to its image input; input precedence requires matching implementation. Interpolator VIDEO output 1 reaches SaveVideo 22. Both saves are active. Keep-audio does not prove bit-perfect sound, VFR correctness or successful frame interpolation.Interpolator requests 2x,motion auto,threshold0.24,keep_audio on; runtime_dir/wine_prefix are empty. Dependencies, actual source-FPS multiplication and hardware support are untested.## Files and selected assets- [Workflow JSON]()- [Selected cover]()- `seedvr2_ema_3b_fp8_e4m3fn.safetensors`## Version, evidence and rightsThis is a local descriptive draft, not a published upgrade. Files01–08 are byte-identical to originals; only the standalone ninth file's historical video-preview field is cleared in its new copy. Originals remain unchanged.No model inference, visual/audio acceptance, wall-clock benchmark or license verification was performed. Covers are selected assets, not accepted render evidence. Weights and external input media are not bundled. Check workflow, nodes, models, LoRAs, source media and cover rights separately; no blanket commercial license is asserted.[Source-provided tutorial](https://www.youtube.com/watch?v=R3W2ab1_TzY) · [Source-provided workflow entry](https://www.runninghub.cn/post/2098776201618878466/?inviteCode=rh-v1111)Links are not live-verified. The workflow is supplied through AI-KSK.

_____________SeedVR2_DLSS___.json

v1.0

<div data-youtube-video="https://youtu.be/R3W2ab1_TzY"><iframe src="https://www.youtube.com/embed/R3W2ab1_TzY" width="853" height="480" allowfullscreen></iframe></div> 
 
AI-KSK ComfyUI workflow: Universal Video Upscale with SeedVR2 and optional DLSS frame interpolation. A standalone post-processing chain for any finished video, independent of the MiniMax H3 generation pack. 
 
Main features: 
- SeedVR2 EMA 3B FP8 video upscaler with selectable target short edge (default 2160p); output short edge = input short edge x multiplier, with block-swap VRAM optimization. 
- DLSS 5 frame interpolation with selectable 2x / 3x / 4x rate after upscaling. 
- Two independent SaveVideo outputs: SeedVR2-only result at the source frame rate, and the final SeedVR2 + DLSS result. 
- No text encoder, VAE, or sampling models required; pure inference-side post processing. 
- Feed it the 768P, 2K, or 4K renders from the MiniMax H3 workflows in the same pack, or any other finished video. 
 
Suggested workflow: 
1. Import the JSON into ComfyUI and install the model files and custom nodes referenced by the workflow. 
2. Provide the required text, image, or video inputs for the selected path. 
3. Adjust resolution, steps, denoise, and prompt controls, then run the connected SaveVideo output. 
 
 
RunningHub workflow: 
https://www.runninghub.ai/zh-cn/post/2098788728777388033?inviteCode=rh-v1111 
 
AI-KSK RunningHub workflow hub: 
https://www.runninghub.ai/user-center/1892566146468511746?inviteCode=rh-v1111 
 
Bilibili tutorial: 
https://www.bilibili.com/video/BV1vFYz6HECU/ 
 
RunningHub registration and AI tools: 
https://www.runninghub.ai/?inviteCode=rh-v1111 
https://www.runninghub.ai/ai-generator/2046514150500524034?inviteCode=rh-v1111 
https://www.runninghub.ai/vip-rights/3?inviteCode=rh-v1111 
https://rhtv.runninghub.ai/projects?inviteCode=rh-v1111 
 
Stable global access: 
https://lelians.com/#/register?code=mFIJ5tZn 
 
TokenRhythm AI model credits: 
https://tokenrhythm.studio/i/rf_tr_kM6wekDquDDZQD4PUNmgiuHA 
 
Youyun Cloud GPU: 
https://passport.compshare.cn/register?referral_code=A7fTQUYY9lCDFJwA6MMvoq 
https://www.notion.so/2b72459a3ac6808f8282f2c687ed8f52 
 
RunPod Cloud GPU: 
https://runpod.io/?ref=sxk5n5m7 
 
Support AI-KSK: 
https://ko-fi.com/aiksk

Universal Video Upscale SeedVR2 + DLSS

<div data-youtube-video="https://youtu.be/R3W2ab1_TzY"></div> 
 
AI-KSK ComfyUI workflow: Universal Video Upscale with SeedVR2 and optional DLSS frame interpolation. A standalone post-processing chain for any finished video, independent of the MiniMax H3 generation pack. 
 
Main features: 
- SeedVR2 EMA 3B FP8 video upscaler with selectable target short edge (default 2160p); output short edge = input short edge x multiplier, with block-swap VRAM optimization. 
- DLSS 5 frame interpolation with selectable 2x / 3x / 4x rate after upscaling. 
- Two independent SaveVideo outputs: SeedVR2-only result at the source frame rate, and the final SeedVR2 + DLSS result. 
- No text encoder, VAE, or sampling models required; pure inference-side post processing. 
- Feed it the 768P, 2K, or 4K renders from the MiniMax H3 workflows in the same pack, or any other finished video. 
 
Suggested workflow: 
1. Import the JSON into ComfyUI and install the model files and custom nodes referenced by the workflow. 
2. Provide the required text, image, or video inputs for the selected path. 
3. Adjust resolution, steps, denoise, and prompt controls, then run the connected SaveVideo output. 
 
 
RunningHub workflow: 
https://www.runninghub.ai/zh-cn/post/2098788728777388033?inviteCode=rh-v1111 
 
AI-KSK RunningHub workflow hub: 
https://www.runninghub.ai/user-center/1892566146468511746?inviteCode=rh-v1111 
 
Bilibili tutorial: 
https://www.bilibili.com/video/BV1vFYz6HECU/ 
 
RunningHub registration and AI tools: 
https://www.runninghub.ai/?inviteCode=rh-v1111 
https://www.runninghub.ai/ai-generator/2046514150500524034?inviteCode=rh-v1111 
https://www.runninghub.ai/vip-rights/3?inviteCode=rh-v1111 
https://rhtv.runninghub.ai/projects?inviteCode=rh-v1111 
 
Stable global access: 
https://lelians.com/#/register?code=mFIJ5tZn 
 
TokenRhythm AI model credits: 
https://tokenrhythm.studio/i/rf_tr_kM6wekDquDDZQD4PUNmgiuHA 
 
Youyun Cloud GPU: 
https://passport.compshare.cn/register?referral_code=A7fTQUYY9lCDFJwA6MMvoq 
https://www.notion.so/2b72459a3ac6808f8282f2c687ed8f52 
 
RunPod Cloud GPU: 
https://runpod.io/?ref=sxk5n5m7 
 
Support AI-KSK: 
https://ko-fi.com/aiksk

Files with SHA256: 467635bc724789df42dae1deaf90dc6bf73df931853041038b872dfa1e785a9a