Multiple Angle Consistent Character with Qwen Edit LoRA

This Qwen Edit workflow lets you AI “edit” an image to show a character in multiple angles. See samples below.

Video with multiple angles.

Software

Workflow overview

This ComfyUI workflow uses Qwen Image Edit, an image generation model that allows you to use a text prompt to carry out image editing, such as adding a person or changing clothes.

To speed up editing, I added Lightx2v’s 4-step LoRA, so that the image editing is four times faster.

The secret sauce of this workflow is the fal’s Qwen Image Edit Multiple Angle LoRA. It is the first multi-angle LoRA trained with 96 different view points of the same object.

Fal has provided a ComfyUI workflow to use the LoRA, but I use VNCCS‘s Visual Camera Control node so that you can define the viewing angle using a graphical dial instead of text prompt.

Qwen Image Edit Multiple Angle workflow

Step 1: Download models

1. Download qwen_image_edit_2511_bf16.safetensors and put it in ComfyUI > models > diffusion_models.
2. Download qwen_2.5_vl_7b_fp8_scaled.safetensors and put it in ComfyUI > models > text_encoders.
3. Download qwen_image_vae.safetensors and put it in ComfyUI > models > vae.
4. Download Qwen-Image-Edit-2511-Lightning-4steps-V1.0-bf16.safetensors and put it in ComfyUI > models > loras.
5. Download qwen-image-edit-2511-multiple-angles-lora.safetensors and put it in ComfyUI > models > loras.

Step 2: Load the workflow

Download the JSON workflow below. Drop the file to ComfyUI to load.

Step 3: Install missing nodes

ComfyUI Manager may not install the missing nodes correctly for this workflow. You will install the ComfyUI_VNCCS_Utils custom nodes.

1. Click Manager.
2. Click Custom Nodes Manager.
3. Search for “VNCCS”
4. Install VNCCS – Collection of utility nodes.

Restart ComfyUI.

Step 4: Upload an image

Upload an image you wish to edit to the Load Image node.

You can use the test image below.

Step 5: Set the camera angle

Set the camera angle by moving the small arrow and dot. The arrow changes lateral view angle. The dot changes the elevation of the camera.

How the node works: The node merely provides a visual dial of the camera angle. It translates the setting into the prompt.

Step 6: Generate images

Click the Run button to run the workflow.

Use cases

1. Camera movement around a character

With consistent character generation in multiple angles, you can use Wan 2.2 First Last Frame workflow to generate a video clip with camera moving around a character.

Using the following images as the first and the last frame of the video.

We can generate a video with revolving camera like this:

2. Consistent character with motions

In this second use case, you use Qwen Image Edit to change the pose of the character.

Use the Wan 2.2 First Last Frame workflow to generate a video clip connecting the two images.

You can also combine the first and the second video to create a longer video.