PlagueKind - LTX2.3 Workflow - I2V - NVFP4 | INT8 Convrot | Ease of use - Eros | Sulphur Compatible
## V6 Notes
- Updated Eros to 1.4 much better.
- You can disable all optional lora apart from distill and eros if you want. The others do enhance the output though.
- Added Volume Boost option.
- Using DMD for audio and video.
- Use 0.715, 0.2412, 0.0 - on 2x FPS sigma if you see jitter
-
## V5 Notes
- Using decoupled distillation for video and audio, and smaller loras. - REQUIRED.
- Updated to Eros 1.3
- FPS bumped to 30fps default.
- 60fps native with 120fps via 2x now compatible.
- Added an actually useable NVFP4 model, even though int8 convrot is more or less the same size and speed.
- You can regain speed by disabling one of the optional loras, start with sulphur.
## V4 Notes
- New reference node injects reference directly into the model, mostly eliminating any drift.
- New DMD distill lora, fixes issues of original distill which broke decompression method and prompt adherence, no need to write a paragraph anymore. Preprocess returns for enhanced movement.
- Use 6-13 steps, default 8-9.
- INT8 Triton VAE for fast decode.
- INT8 Convrot added for increased speed on all cards.
- JoyAI lora another lora similar to omninft, this helps adherence, dialog, interaction and more.
- If using the nvfp4 transformer make sure to describe anatomy that isn't already visible in the image. I don't recommend using the nvfp4 text encoder for realism.
---
## V3 Notes
TURN DOWN NEW EROS LORA TO 0.7 I SET IT AT 0.9 and it doesn't look correct like that.
- Added Prompt Relay Node for advanced timeline prompting while keeping simplicity.
- Added VHS save video. AV1/WEBM CRF0 produces virtually lossless video with extremely small file size.
- Updated Eros Lora.
- Moved RTX VSR into sampler graph.
## V2 Notes
- Upscale and 2x fps have added tiled sampler to improve speed. if you don't like the output, experiment with cfg pp sampler, or add 0.715 to sigma. Current sigma are done on purpose to retain identity and quality of first pass.
## V1 Notes
- Focused Mainly on Speed and ease of usability while retaining single pass quality and likeness.
- First pass is more or less optimized.
- Identity Retention is pretty good.
- First pass balanced for speed and quality.
- Upscale and 2x fps work but will be optimized later.
- I would avoid using RTX VSR with the whole chain active unless you have like 96gb ram.
- Nvfp4 models are only for Blackwell, using them will not save you speed if you don't have Blackwell, but they should still work.
- Default settings of the first pass should produce good quality and sharp. I personally like to use this for a single pass and just render straight at 1280.
- Bumped FPS to 30 because it looks way bettwen imo, quality seems clearer overall and it gets rid of that ltx base fps look.
---
## Memory Chunk Note
- This saves vram and speeds up it/s especially on longer videos
- With 16gb of VRAM 3 chunks seems optimal for me
- If you get triton errors, enter the model loader subgraph and disable use triton kernels
---
## Requirements
- [ComfyUI-PlagueKind-Nodes](https://github.com/PlagueKind/Comfyui-PlagueKind-Nodes)
- [ComfyUI-KJNodes](https://github.com/kijai/ComfyUI-KJNodes)
- [Rgthree-comfy](https://github.com/rgthree/rgthree-comfy)
- [10s-nodes](https://github.com/TenStrip/10S-Comfy-nodes)
- [Nvidia_RTX_Nodes_ComfyUI](https://github.com/Comfy-Org/Nvidia_RTX_Nodes_ComfyUI)
- [ComfyUI-VideoHelperSuite](https://github.com/Kosinkadink/ComfyUI-VideoHelperSuite)
- [ComfyUI-PromptRelay](https://github.com/kijai/ComfyUI-PromptRelay)
- [Sage Attention](https://github.com/thu-ml/SageAttention) (optional)
---
### Model Choices
[ltx int8 convrot model](https://huggingface.co/Kijai/LTX2.3_comfy/resolve/main/diffusion_models/ltx-2.3-22b-dev_transformer_only_int8_convrot.safetensors)
[ltx nvfp4 model](https://huggingface.co/Bedovyy/LTX2.3_transformer_only_comfy/resolve/main/ltx-2.3-22b-dev_transformer_only_nvfp4mixed_input_scaled.safetensors)
### Latent Upscalers
[ltx-2.3-spatial-upscaler-x2-1.1.safetensors](https://huggingface.co/Lightricks/LTX-2.3/resolve/main/ltx-2.3-spatial-upscaler-x2-1.1.safetensors?download=true)
[ltx-2.3-temporal-upscaler-x2-1.0.safetensors](https://huggingface.co/Lightricks/LTX-2.3/resolve/main/ltx-2.3-temporal-upscaler-x2-1.0.safetensors?download=true)
### VAE
[LTX23_audio_vae_bf16.safetensors](https://huggingface.co/Kijai/LTX2.3_comfy/resolve/main/vae/LTX23_audio_vae_bf16.safetensors?download=true)
[LTX23_video_vae_bf16.safetensors](https://huggingface.co/Kijai/LTX2.3_comfy/resolve/main/vae/LTX23_video_vae_bf16.safetensors?download=true)
### Dual CLIP (Encoder and Projection) Choices
[gemma_3_12B_it_fp4_mixed.safetensors](https://huggingface.co/Comfy-Org/ltx-2/resolve/main/split_files/text_encoders/gemma_3_12B_it_fp4_mixed.safetensors?download=true)
[gemma_3_12B_it_fp8_scaled.safetensors](https://huggingface.co/Comfy-Org/ltx-2/resolve/main/split_files/text_encoders/gemma_3_12B_it_fp8_scaled.safetensors?download=true)
[ltx-2.3_text_projection_fp8.safetensors](https://huggingface.co/Winnougan/LTX-2.3-INT8/resolve/main/ltx-2.3_text_projection_fp8.safetensors?download=true)
[ltx-2.3_text_projection_bf16.safetensors](https://huggingface.co/Kijai/LTX2.3_comfy/resolve/main/text_encoders/ltx-2.3_text_projection_bf16.safetensors?download=true)
### Distilled LoRA
[LTX2.3_DMD_fro99-avgrank47.safetensors](https://huggingface.co/maximsobolev275/LTX2.3_DMD-10eros-resized/resolve/main/LTX2.3_DMD_fro99-avgrank47.safetensors)
---
### Optional LoRA
[LTX-2.3-OmniNFT-RL-Lora_bf16.safetensors](https://huggingface.co/Kijai/LTX2.3_comfy/resolve/main/loras/LTX-2.3-OmniNFT-RL-Lora_bf16.safetensors?download=true)
[LTX_10Eros-v13_LoRA_fro99-avgrank97.safetensors](https://huggingface.co/maximsobolev275/LTX-10Eros-LoRA-r768/resolve/main/LTX_10Eros-v13_LoRA_WAN2GP_fro99-avgrank97.safetensors)
[JoyAI-Echo-content_r256.safetensors](https://huggingface.co/TenStrip/LTX2.3_JoyAI_Lora_Extracted/resolve/main/JoyAI-Echo-content_r256.safetensors)
[LTX_10Eros-v14_LoRA_fro99-avgrank104.safetensors](https://huggingface.co/maximsobolev275/LTX-10Eros-LoRA-r768/resolve/main/LTX_10Eros-v14_LoRA_fro99-avgrank104.safetensors)
### Optional Model converted by me
[10Eros_v1_experimental_nvfp4.safetensors](https://huggingface.co/Plaguekind/10Eros_v1_experimental_nvfp4/resolve/main/10Eros_v1_experimental_nvfp4.safetensors?download=true)
If you find this project useful and want to support development:
Monero (XMR): 865BrcfWLdwELwuq5faV1uVTbh93zVK6AUYLY2c3mX6sFfAGRfS6axe1kBTYYKuM7ccN7zBZDAZvnT7E4NKmUazySdbpc7p
Thank you for your support.
Description
- Upscale and 2x fps have added tiled sampler to improve speed. if you don't like the output, experiment with cfg pp sampler, or add 0.715 to sigma. Current sigma are done on purpose to retain identity and quality of first pass.
FAQ
Comments (9)
Hi, II've tried your WF and have immediately this error (All is correctly installed and I've tried with the two models of safetensors, nothing to do..): RuntimeError("ERROR: VAE is invalid: None\n\nIf the VAE is from a checkpoint loader node your checkpoint does not contain a valid VAE.")
RuntimeError: ERROR: VAE is invalid: None
If the VAE is from a checkpoint loader node your checkpoint does not contain a valid VAE.
This workflow is setup for transformer only builds.
Thanks for the reply, I actually solved it myself, the safetensors models are not the problem, there are two errors in workflow vae nodes, now it works, but only without the loras node, although installed correctly when it is enabled, it blocks the entire workflow, and without loras the result unfortunately is not very good.
@omadastered youll need to explain your issue a bit more because i just asked on discord and no one has ever encountered this problem. Myself included. Make sure your venv is up to date and clean, as well as being on the newest version of comfyui
This is the best workflow i ever used on LTX. Insane sharpness and quality. I reworked it a bit by adding Prompt Relay Encode node (for better following the prompt). Results are amazing! Thank you for sharing it!
Thanks alot for the comment. I'll try that out and consider it for v3 thanks
@Plaguekind LTX Director 2.0 is crazy. Thats even better then relay!
@alessandrowww Yeah It's pretty awesome, but I want to keep the workflow clean and somewhat simplistic. Maybe in the future I'll make a director version.
@Plaguekind I would appreciate! I tried your wf its very nice, but director is a node i can't ditch^^