README
ComfyUI Workflows
A personal collection of ComfyUI workflows for AI video generation, upscaling, and video editing. Each workflow entry includes a description, example output, requirements, and usage notes.
Drag-and-drop any .json file directly into ComfyUI to load the workflow.
Posts
| Post | Date |
|---|---|
| MOMENTS — Making of | 2026-03-30 |
MiniMax H3
| Workflow | Description | Type | Updated |
|---|---|---|---|
| CrossView Warp | V2V new-camera-angle LoRA — MoGe-2 geometry warp as an aligned guide, source clip as a Ref2VA reference, two-pass with audio (step 3.5k) | Video gen | 2026-09-20 |
LTX Video 2.3
| Workflow | Description | Type | Updated |
|---|---|---|---|
| IC-LoRA CrossView Warp (v2) | V2V new-camera-angle IC-LoRA — MoGe-2 geometry warp, keyframed virtual camera path set in the node's live preview (v2, step 6k) | Video gen | 2026-08-15 |
| IC-LoRA CrossView Warp (v0.9) | V2V new-camera-angle IC-LoRA — set the virtual second camera by numbers (azimuth/elevation/distance) via a depth-warp node (PoC, step 18k) | Video gen | 2026-07-24 |
| IC-LoRA CrossView Prompt (v0.9) | V2V new-camera-angle IC-LoRA — re-render a clip from a prompt-described virtual second camera (pilot, step 13.7k) | Video gen | 2026-07-11 |
| IC-LoRA Cameraman v2 | I2V with IC-LoRA camera motion guidance (Cameraman v2, step 14k) — dev + distilled LoRA + 2x upscaler | Video gen | 2026-06-24 |
| Arcane Jinx Character LoRA | T2V with audio using a custom-trained Arcane Jinx character LoRA on LTX 2.3 — 22B | Video gen | 2026-05-06 |
| ReStyle IC-LoRA | V2V style transfer using a custom-trained ReStyle IC-LoRA on LTX 2.3 — 22B | Video edit | 2026-05-05 |
| IC-LoRA Cameraman v1 | I2V with IC-LoRA camera motion guidance from a reference video clip | Video gen | 2026-04-07 |
| I2V Two-Pass | Two-pass I2V with start/end guide images, 720p→1080p upscale, and audio support | Video gen | 2026-03-30 |
WAN 2.2
| Workflow | Description | Type | Updated |
|---|---|---|---|
| Face Detailer | SAM2-based segmentation and WAN 2.2 detail pass on existing video | Video edit | 2026-03-29 |
| VACE Endless Extension | Extend a video indefinitely using VACE looping | Video gen | 2025-11-10 |
| Upscaling v1 | Per-subject crop upscaling with VFI frame interpolation | Upscaling | 2025-10-13 |
| Lightning GGUF | Fast T2V with cached T5 encoding and Lightning-tuned steps | Video gen | 2025-08-09 |
| T2V A14B | Text-to-video with the WAN 2.2 A14B model, GGUF quantized | Video gen | 2025-07-30 |
| VACE | Video-to-video with automatic subject segmentation and depth conditioning | Video edit | 2025-07-15 |
HunyuanVideo
| Workflow | Description | Type | Updated |
|---|---|---|---|
| RF-Inversion | Video editing via RF-inversion — relight, restyle, or modify existing video | Video edit | 2025-01-23 |
Z-Image
| Workflow | Description | Type | Updated |
|---|---|---|---|
| Two-Stage | Two-pass T2I workflow with refinement and color grading | Image gen | 2026-03-29 |
Support
Producing and sharing this kind of open-source work requires renting cloud GPUs, which gets expensive quickly. If you find it useful and would like me to keep contributing, your support is very much appreciated:
Contributors
Cseti
28 commits
CO
Claude Opus 4.8
1 commits
Similar repos
README
ComfyUI Workflows
A personal collection of ComfyUI workflows for AI video generation, upscaling, and video editing. Each workflow entry includes a description, example output, requirements, and usage notes.
Drag-and-drop any .json file directly into ComfyUI to load the workflow.
Posts
| Post | Date |
|---|---|
| MOMENTS — Making of | 2026-03-30 |
MiniMax H3
| Workflow | Description | Type | Updated |
|---|---|---|---|
| CrossView Warp | V2V new-camera-angle LoRA — MoGe-2 geometry warp as an aligned guide, source clip as a Ref2VA reference, two-pass with audio (step 3.5k) | Video gen | 2026-09-20 |
LTX Video 2.3
| Workflow | Description | Type | Updated |
|---|---|---|---|
| IC-LoRA CrossView Warp (v2) | V2V new-camera-angle IC-LoRA — MoGe-2 geometry warp, keyframed virtual camera path set in the node's live preview (v2, step 6k) | Video gen | 2026-08-15 |
| IC-LoRA CrossView Warp (v0.9) | V2V new-camera-angle IC-LoRA — set the virtual second camera by numbers (azimuth/elevation/distance) via a depth-warp node (PoC, step 18k) | Video gen | 2026-07-24 |
| IC-LoRA CrossView Prompt (v0.9) | V2V new-camera-angle IC-LoRA — re-render a clip from a prompt-described virtual second camera (pilot, step 13.7k) | Video gen | 2026-07-11 |
| IC-LoRA Cameraman v2 | I2V with IC-LoRA camera motion guidance (Cameraman v2, step 14k) — dev + distilled LoRA + 2x upscaler | Video gen | 2026-06-24 |
| Arcane Jinx Character LoRA | T2V with audio using a custom-trained Arcane Jinx character LoRA on LTX 2.3 — 22B | Video gen | 2026-05-06 |
| ReStyle IC-LoRA | V2V style transfer using a custom-trained ReStyle IC-LoRA on LTX 2.3 — 22B | Video edit | 2026-05-05 |
| IC-LoRA Cameraman v1 | I2V with IC-LoRA camera motion guidance from a reference video clip | Video gen | 2026-04-07 |
| I2V Two-Pass | Two-pass I2V with start/end guide images, 720p→1080p upscale, and audio support | Video gen | 2026-03-30 |
WAN 2.2
| Workflow | Description | Type | Updated |
|---|---|---|---|
| Face Detailer | SAM2-based segmentation and WAN 2.2 detail pass on existing video | Video edit | 2026-03-29 |
| VACE Endless Extension | Extend a video indefinitely using VACE looping | Video gen | 2025-11-10 |
| Upscaling v1 | Per-subject crop upscaling with VFI frame interpolation | Upscaling | 2025-10-13 |
| Lightning GGUF | Fast T2V with cached T5 encoding and Lightning-tuned steps | Video gen | 2025-08-09 |
| T2V A14B | Text-to-video with the WAN 2.2 A14B model, GGUF quantized | Video gen | 2025-07-30 |
| VACE | Video-to-video with automatic subject segmentation and depth conditioning | Video edit | 2025-07-15 |
HunyuanVideo
| Workflow | Description | Type | Updated |
|---|---|---|---|
| RF-Inversion | Video editing via RF-inversion — relight, restyle, or modify existing video | Video edit | 2025-01-23 |
Z-Image
| Workflow | Description | Type | Updated |
|---|---|---|---|
| Two-Stage | Two-pass T2I workflow with refinement and color grading | Image gen | 2026-03-29 |
Support
Producing and sharing this kind of open-source work requires renting cloud GPUs, which gets expensive quickly. If you find it useful and would like me to keep contributing, your support is very much appreciated:
Similar repos
Contributors
Cseti
28 commits
CO
Claude Opus 4.8
1 commits