Enhancements to text2video

[iChristGit]

Is it possible

-Support the default ltx2 model and workflow

-Allow to upload an input image or normal text prompt?
So user can choose if img2vid or txt2vid

That would be very good!

[Haervwe]

Yes it's possible , lemme update the tool :)

[iChristGit]

Thank you! I love using all of your tools!!

[iChristGit]

Any roadblocks with enabling ltx2? I really want to incorporate text2vid and imag2vid in my open-webui

[Haervwe]

I'm trying to get it to work! My GPU can't handle it it seems, but we can if you like coordinate and go on discord or asynchronously test it out, also the tools have several refactor due to file management that were in the way, but I have not forgotten!

[iChristGit]

I can definitely help you test newer version, my PC can handle LTX2 with offloading of the LLM, can you share it directly here or maybe update the tool and il see how it goes

[iChristGit]

I created the new LTX2.3 tool!
https://openwebui.com/posts/ltx_23_video_generator_text_to_video_image_to_vide_d31d1572

img2video, txt2video, automatically choose the video resolution depending on initial image input, choosing 5s 10s 15s etc, embedded player.

If you can run this workflow, this is pretty good!

LTX_2.3_t2v_00012_.mp4