thanks for sharing , if you help , it will be fantastic
Hi Solricks , thanks for sharing the worflow for ltx2.5, i wanted to use the add audio option evrytime it stuck and upscale ( 3 steps) , i removed it and it got stuck anyway. Is it on my end? i can run the commfyui ltx2.5 just fine , any help when you have time is appreciated. CHEERS
Hi! Thanks for trying the workflow.
It’s working fine on our end with an RTX 3090, RTX 5090 and Blackwell 6000
Have you tried lowering or adjusting the resolution and video duration before running it? LTX 2.5 can be quite demanding depending on those settings.
Also, which GPU are you using and how much VRAM does it have?
If you share your resolution, duration, and GPU model, I can take a look and try to figure out what might be causing it.
hi. thanks for th qwick answer , i have a :
[INFO] Total VRAM 16380 MB, total RAM 65372 MB
[INFO] pytorch version: 2.9.1+cu130
[INFO] Set vram state to: NORMAL_VRAM
[INFO] Device: cuda:0 NVIDIA GeForce RTX 4060 Ti : cudaMallocAsync
I use 0.6 mega pixel resolution with ltx2.5 sometimes higher it works fine add with minimax h3 ill go 0.9 even
screenchot of the resolution i use
Thanks in advance.
Hi,
Your screenshot is actually set to 0.9 MP, not 0.6 MP — the Resolution Selector is currently showing 0.9 MP.
At a 9:16 aspect ratio, 0.9 MP is already quite demanding for LTX-2.5 on an RTX 4060 Ti with 16 GB VRAM.
On top of that, the second-stage 2x spatial upscale/refine in this workflow increases peak memory usage even further. The official LTX-2.5 two-stage T2V/I2V workflows also use a separate spatial upscaler, so additional VRAM usage during the upscale stage is expected.
I would first try lowering the resolution to around 0.4–0.6 MP and see if the workflow completes successfully.
Hi , actually i didn't touch it was set too jhih so a put 0.4 and its working now event abote 10 sec ( it might be do to latest update of comfui too )
thanks for you help
try megapixels 0.4 and multiple 64