Open Source Video Showdown - LTX, CogX, HunYuan Text to Video

preview_player
Показать описание

Links to workflows below description!

In this episode, we dive into a comprehensive comparison of three leading open-source AI video generation tools: CogXVideo, LTXVideo, and Hun Yuan.

Using familiar prompts from the newly released Sora videos, we evaluate the performance, speed, and quality of each tool.

Discover the unique strengths and weaknesses of each model and see the remarkable results they can produce.

Thanks to our sponsor, Mimic PC, we were able to execute this comparison efficiently. Watch now to find out which AI video generator comes out on top!

00:00 Introduction to Sora's Release and Video AI Alternatives
00:26 Overview of the Top Open Source Video AI Contenders
00:46 Setting Up the Comparison Test
01:10 Using Mimic PC for Efficient Video Rendering
01:41 Rendering Times and Quality Trade-offs
01:57 Demonstrating Comfy UI Workflows
02:31 First Prompt: Spaceman Movie Trailer
04:13 Second Prompt: Skateboarding Border Collie
06:21 Third Prompt: Pirate Ships in a Coffee Cup
07:35 Fourth Prompt: Zen Garden in a Glass Sphere
09:51 Fifth Prompt: Glass Tortoise on a Black Sand Beach
10:33 Sixth Prompt: Waves Crashing
11:23 Seventh Prompt: Woolly Mammoths in Snow
12:39 Eighth Prompt: Jellyfish in a Cityscape
13:22 Ninth Prompt: Stylish Woman in Tokyo
14:52 Tenth Prompt: Product Reviewer in a Studio
15:27 Conclusion and Final Thoughts

👍 LIKE If you found this video valuable. 🙂
🥰 SHARE If you know someone who might enjoy this video.
⏬ DOWNLOAD or ADD This video to your PLAYLIST for easy access later.
💬 COMMENT Your thoughts and questions are welcome!

That's what keeps me going!

🔔 ​​And make sure to hit the NOTIFICATION BELL to stay updated! 🔔

🌎EXPLORE

🗓️ MY HISTORY
🎙️ Voice Over Artist 30+ years
📻 Broadcaster / Actor / Creative

❓ASK ME

📌 FOLLOW ME

🪙SUPPORT
If you want to support me, the best thing to do is to share the content… sharing is caring!
If you want to support me financially (completely optional and voluntary, but a lot of people have asked for this):

Рекомендации по теме
Комментарии
Автор

🌀✨Hey! Quick favor... ✨🌀

If you found value in this video, a simple LIKE 👍 and hitting SUBSCRIBE 🛎 helps more than you know. It lets YouTube know we’re onto something good here and helps others discover it too. ⚛

I’d love to hear what you think! 💬 Drop your thoughts, insights, or questions in the comments below. 👇

Your support means everything. Let’s keep exploring the creative uses of AI together!

#BobDoyleMedia #LikeAndSubscribe #YourSupportMatters #ThankYou

BobDoyleMedia
Автор

Four things I really like about Bob's videos:
1) the content
2) his deadpan humor
3) He doesn't say "go ahead and" approximately 178 times in 10 minutes.
4) He doesn't say "let's dive in" before getting to the freaking point that his audience came for and doesn't need to be told is coming.
Bravo, sir.

peterjarneson
Автор

The Hunyuan one is pretty cool. i been playing around with it and it's first time ever in a video generation tool locally on my PC and actually got results that are impressive. Only sad thing is how incredibly short the videos are, on a 4090 even if you go down to a super low resolution of like 512 by 320 it still can only do like 8 seconds max, and doing 720p your only getting like 2 seconds. Still fun to play with. i don't know if locally we'lll ever get to make cool videos of any decent length, not unless by some miracle graphics cards with 100gb come out and are affordable.

zengrath
Автор

ty Bob and ty china for this hunyuan model

p_p
Автор

How much time does it take to generate 5 sec video using H100 80gb with hunyuan? Seems slower even with 80 gb vram

abuzar
Автор

Does anyone have github links for theae?

luismiguelable
Автор

Is there any model that you can train your own datasets ? Not images, i am talking about videos.

rrlaishram
Автор

God bless Kijai. I really don't think any of them would be as popular if not for him. But to be fair I don't think we're giving them the full weigh in if the full models require 24GB-80GB VRAM (which we know OpenAI has). I don't think MimicPC is sitting on a stack of H200s. Plus all their templates run Kijai. Would you be willing to do one for maybe Genmo Mochi-1, and Hunyuan on a beast of a server? If MimicPC doesn't have one I think VastAI does. I just want to see it fully once.

RamonRodgers
Автор

I really need help. I’m trying to put my voice into my song on Suno. Anyone can help me please. Thanks

KaioshinAemi
Автор

You missed good contender pyramid Flow 😮

FSK
Автор

V2V with HunYuan is pretty sweet. I'm throwing barebones Blender and iClone scene mockups at it.

BlackMattk
Автор

Hi. Is there a version of hunyuan image to video?

onlineispections
Автор

You need to attempt these Prompts with Sora! For all we know they prompted 500 times before they got the outputs they showed the World.

EyeMenu
Автор

Don't forget mochi on comfyui and LTX with img to video 😀

MOSTRAAKX
Автор

I am not spending a cent until I find the ONE

todaychange-
Автор

I want to own one I want to Hug one you should do tosh.0 style video comedy ....that one had me laughing my ass off

friedguyfriedguy
Автор

hun-waan its not that hard to say. also your video is out of sync with the audio

RandPrint
Автор

Why are your video's always out of sync?

synriser