Wan 2.2 - All in Wan I2V (I2V, F2LF, SVI with optional F2LF, NAG, LTX for V2A, Pulse of Motion, Lora Optimizer, CFG-Ctrl, 4 modes, ...) — v2.1

Use this Wan 2.2 workflow to turn images and prompts into video, make a clip with a chosen end frame, extend video with SVI 2.0 Pro, or add audio to existing or newly generated video with LTX-2.3. Inputs include images, prompts, optional end frames, or video; outputs are video or

Choose regular image-to-video, F2LF with a chosen end frame, video extension with SVI 2.0 Pro, or audio addition with LTX-2.3.

For image-to-video, use a starting image and prompt; F2LF also uses a chosen end frame.

SVI 2.0 Pro extends a video, while LTX-2.3 adds audio to an existing or newly generated video.

This v2.1 workflow uses Wan Video 2.2 I2V-A14B as its base model.

ComfyUI node packs named for the workflow include ComfyUI-Impact-Pack, ComfyUI-KJNodes, ComfyUI-Easy-Use, ComfyUI_Fill-Nodes, ComfyUI-VideoHelperSuite, and ComfyUI-Custom-Scripts.

Model files named for the workflow include gemma_3_12B_it_fp8_scaled.safetensors, ltx-2.3-22b-dev-fp8.safetensors, ltx-2.3-22b-distilled-lora-384.safetensors, SVI_v2_PRO_Wan2.2-I2V-A14B_HIGH_lora_rank_128_fp16.safetensors, SVI_v2_PRO_Wan2.2-I2V-A14B_LOW_lora_rank_128_fp16.safet,

Other named model files include wan2.2_i2v_high_noise_14B_fp16.safetensors, wan2.2_i2v_low_noise_14B_fp16.safetensors, wan2.2_i2v_lightx2v_4steps_lora_v1_high_noise.safetensors, wan2.2_i2v_lightx2v_4steps_lora_v1_low_noise.safetensors, Wan22_I2V_VBVR_HIGH_rank_64_fp16.safetensors

Pulse of Motion can adjust video speed automatically; manual FPS control is also available.

NAG supports a negative prompt with CFG 1.0; CFG-Ctrl/SMC-CFG is another prompt-following option.

The LoRA optimizer is a separate option from the regular loader for combining multiple LoRAs, and is off by default.

If CFG-Ctrl is active, the author advises against using lightx2v on the high-noise path.

Suggestion · non vérifié