FLUX2 memory with and without images
- Dominant language
- Python
- Stars
- 133k
- Forks
- 15.7k
- Avg merge
- 1d 6h
- Merged PRs (30d)
- 155
Description
### Custom Node Testing
- [ ] I have tried disabling custom nodes and the issue persists (see [how to disable custom nodes](https://docs.comfy.org/troubleshooting/custom-node-issues#step-1%3A-test-with-all-custom-nodes-disabled) if you need help)
### Expected Behavior
less ram usage while no use of input images
### Actual Behavior
more ram usage while using no input images
### Steps to Reproduce
use the actual workflow for flux2 from comfy
use a gguf model Q4
attach input images - watch vram usage ~14-15GB
disable images vram usage >16GB
### Debug Logs
```powershell
no logs, is no error message
```
### Other
_No response_
Contributor guide
Research direction
Start by reproducing the actual FLUX2 ComfyUI workflow with a GGUF Q4 model, comparing VRAM usage with input images attached and disabled. Check whether the difference is reproducible without custom nodes; done should identify the cause of the higher no-image usage and confirm a measured reduction without changing expected workflow behavior.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python, pytorch
- Domain
- machine-learning, performance
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 30/100