10Eros LTX 2.3 image-to-video
huggingface space using comfyui backend for 10eros LTX 2.3 fp8 mixed checkpoint for I2V with native audio. make sure to upload a starting image first, write a prompt, optionally pick a preset, press enhance prompt to expand a short concept into a detailed video prompt tuned specifically for LTX. native audio is generated jointly with video. If your generations get limited by ZeroGPU duration, feel free to modify the ZeroGPU budget section.
preset
1 41
0 1
0 1
0 1
0 1
0 1
0 1
0 1
0 2
0 2
0 1
0 1
512 1536
512 1536
0 1
0 1
0 0.5
0 1
0 2
2 4
0 24
0.25 0.5
0 10
controls how each lora affects the audio stream (loras default to applying equally to video + audio). set to 0 to stop a lora from influencing audio while keeping its video effect.
0 1
0 1
0 1
0 1
0 1
0 1
0 1
0 2
0 2
0 1
0 1
0 1
0 12
10 2000
0 1
refine sigmas (original = workflow default; tuned = drops 0.715 high step, preserves identity through refine; custom = your own)
20 540
0 540