LTX2.3 GGUF Simple i2v + Last Frame 16GB/low vram — v3 — v5

This LTX 2.3 ComfyUI workflow is for image-to-video creation: it uses an image and prompt, includes last-frame and reference-audio nodes, and assembles a video.

The graph includes latent upscaling, RIFE VFI, video combining, and image saving nodes.

The graph has LoadImage and LoadAudio inputs, LTXVReferenceAudio and CLIPTextEncode nodes, plus VHS_VideoCombine and SaveImage output nodes.

The description offers two diffuser choices: 10Eros_v1.4-Q4_K_S.gguf or ltx-2.3-22b-dev-Q4_K_S.gguf, and says to use both listed text encoders.

The workflow lists these model files: 10Eros_v1.4-Q4_K_S.gguf, gemma_3_12B_it_fp8_scaled.safetensors, lora_weights.safetensors, ltx-2-spatial-upscaler-x2-1.0.safetensors, ltx-2.3-22b-dev_embeddings_connectors.safetensors, LTX23_audio_vae_bf16.safetensors, LTX23_video_vae_bf16.s健

It also lists rife49.pth and taeltx2_3.safetensors, plus the node packs comfyui-kjnodes, comfyui-videohelpersuite, rgthree-comfy, and was-ns.

The description names a 1.1 latent upscaler and says a distilled LoRA is required, while the workflow lists a 1.0 upscaler and no distilled LoRA; verify the matching files before setup.

Suggestion · non vérifié

The description says comfyui_smznodes can break LTXV workflows; check extension compatibility if you use it.

Suggestion · non vérifié