Type in next scene prompt and/or multiple angles prompt. Run.
Runtime profile
Source description
Howto
Type in next scene prompt and/or multiple angles prompt. Run.
Description
This workflow allows you to develop an image into another scene by utilizing the next scene lora and the multiple angles lora. Which can be found here:
https://huggingface.co/lovis93/next-scene-qwen-image-lora-2509/tree/main
AI-generated commentary
AI-generated explanation based on source and configuration details. Suggestions are clearly labeled.
Use this workflow to develop an image into another scene or camera angle by entering a next-scene prompt, a multiple-angles prompt, or both; the output is the developed image.
Enter a next-scene prompt and/or multiple-angles prompt, then run the workflow.
Provide an image and prompt text; the output is the image developed into the new scene or angle.
It works with Qwen Image Edit 2509 and the Lightning LoRA; its base model is Qwen.
The workflow requires the rgthree-comfy node pack.
Listed model files are qwen_image_edit_2509_fp8_e4m3fn.safetensors, qwen_2.5_vl_7b_fp8_scaled.safetensors, and qwen_image_vae.safetensors.
Use a next-scene prompt for the new scene, a multiple-angles prompt for camera-angle changes, or both.
Suggestion · not verified
The workflow uses a next-scene LoRA and a multiple-angles LoRA for these edits.
Before running, verify that qwen_2.5_vl_7b_fp8_scaled.safetensors is available for your text-encoder setup.
Suggestion · not verified
The stated estimate is 8 GB of VRAM, with the amount depending on image size.
Does this need to be edited?
Sign in to send an edit request.
Sources
1 sourceSource excerpts
1 excerptSource context: 790 downloads · Type Workflows · Base model Qwen
It works with Qwen Image Edit 2509 and the Lightning lora to speed things up.
Requirements
It depends of the size of the images. But 8 Gb Vram should be enough.
Estimated VRAM requirement
Estimate unavailable
19.0 GB across 1 of 3 model files. Model file total + 25% loading overhead + 2 GB execution buffer, rounded up.
Requirements
4 requirementsqwen_2.5_vl_7b_fp8_scaled.safetensors
ConflictText encoder · Unknown
Possible sources
Checkpoint · 19.0 GB · SAFETENSORS · fp8 · Unknown
VAE · Unknown
Node pack · Registry