LTX-2.3 I2V AudioSync Simple Workflow V4b (RTX-VSR)

LTX-2.3 I2V AudioSync Simple Workflow V4b (RTX-VSR) is for turning one image and audio such as an MP3 into an audio-synchronized video with RTX Video Super Resolution upscaling.

It includes `Clear Cache All` and `Clean VRAM Used` nodes.

`TextGenerateLTX2Prompt` can analyze the image and enhance the prompt.

Provide one image and audio such as an MP3; the output is an audio-synchronized video.

The setup notes require ComfyUI 0.16.x and SageAttention.

Keep `ComfyUI-LTXVideo` off PyTorch `2.12.0+cu130`; the notes also flag ComfyUI `0.17.x` and later as unstable.

Suggestion · not verified

The workflow notes list these model files: `ltx-2.3-22b-dev-fp8.safetensors`, `gemma_3_12B_it_fp4_mixed.safetensors`, `ltx-2.3-22b-distilled-lora-384.safetensors`, and `ltx-2.3-spatial-upscaler-x2-1.0.safetensors`.

Additional listed model files are `ltx-2.3-22b-distilled-lora-384-1.1.safetensors`, `ltx-2.3-spatial-upscaler-x2-1.1.safetensors`, `gemma-3-12b-it-abliterated_heretic_lora_rank64_bf16.safetensors`, and `gemma-3-12b-it-abliterated_lora_rank64_bf16.safetensors`.

Listed node packs include `comfyui_nvidia_rtx_nodes`, `comfyui-easy-use`, `comfyui-kjnodes`, `ComfyUI-LTXVideo`, `comfyui-videohelpersuite`, and `rgthree-comfy`.

The initial RTX Video Super Resolution scale is `2`.

Enter `yes` to enable prompt enhancement or `no` to bypass it; if `TextGenerateLTX2Prompt` refuses, use `no`.

Suggestion · not verified

Before setup, verify that `comfyui_nvidia_rtx_nodes` and `gemma-3-12b-it-abliterated_heretic_lora_rank64_bf16.safetensors` are available.

Suggestion · not verified

Audio longer than 60 seconds may distort the image, and 2D anime-style images may also distort.