Ovi (only 10 Steps with turbo lora) Open-source AI video with audio! Veo 3.1 In ComfyUI Locally

Use this ComfyUI workflow to create AI video with audio, including talking-character video, from an image or text input; it supports image-to-video and text-to-video.

Choose the image-to-video path for an image input or the text-to-video path for a text input.

Inputs are an image for image-to-video or text for text-to-video.

The workflow is described as creating AI video with audio, including talking-character video.

It is described for local use in ComfyUI, with Wan Video 2.2 TI2V-5B as the base model.

Listed model files include mmaudio_vae_16k_bf16.safetensors, mmaudio_vocoder_bigvgan_best_netG_bf16.safetensors, taew2_2.safetensors, and Wan2_2_Ovi_Audio_fp8_e4m3fn_scaled_KJ.safetensors.

Additional model files are Wan2_2_Ovi_Video_fp8_e4m3fn_scaled_KJ.safetensors, Wan2.2_VAE_bf16.safetensors, and Wan22_TI2V_5B_Turbo_lora_rank_64_fp16.safetensors.

Listed node packs are comfyui-kjnodes, comfyui-videohelpersuite, and ComfyUI-WanVideoWrapper.

It is presented as a 10-step workflow using a turbo LoRA.

Before use, verify that the listed model files and node packs are available in your ComfyUI setup.

Suggestion · not verified

No license is specified in the listing, so verify applicable usage terms before redistributing the workflow.

Suggestion · not verified