Stable Diffusion 3: Open Weights for SD-3 is now Available! The Best Text-To-Image Opensource Model

In this video, I'll be talking about the recently released Stable Diffusion 3 model. This model throws every other model like Midjourney into the water. This is an opensource model. We'll be checking it out in this video. You can easily use this Text-To-Image model to replace Midjourney and stop paying for that.

---------

Key Takeaways:

📢 Big News: Stable Diffusion 3 Weights Released! Stable Diffusion has now released the open weights for Stable Diffusion 3, allowing you to use it locally. 🚀

💻 Run on Consumer PCs and Laptops! This smaller model is perfect for running on simple GPUs and laptops, making advanced AI image generation accessible to everyone. 🖥️

🎨 Text-to-Image Capabilities Enhanced! One of the coolest features is its ability to generate images with perfect text, solving a major problem in AI-generated art. ✍️

🔍 Photorealism and Efficiency Boosted! The model is more photorealistic, addresses common AI image pitfalls, and is resource-efficient. 📸

📜 New License Restrictions! The model is available under the Stability Non-Commercial Research Community License, limiting commercial use. 📄

🖼️ Try It on Hugging Face! You can try the model on Hugging Face without downloading it. Experience the new features and see how it performs! 🌐

💔 Concerns Over Closed-Source Shift! There's growing concern that Stability AI is moving towards a closed-source approach, limiting the community's use. 🚧

---------

Timestamps:

00:00 - Introduction
00:06 - About Stable Diffusion 3's New Release
01:55 - Stable Diffusion 3 Features
02:58 - Non-Commercial License of SD-3
04:49 - Demo & Image Generations
08:03 - Ending

[alekjwrgnwekfgn]

“Will Smith eating noodles. “. I fell out of my chair. Haven’t laughed like that in a long time.

[NLPprompter]

will Smith should open noodles and spaghetti chains on this point

[munirable]

can you show how to install on linux with cpu?

[talebelm]

Hey, have a question ( it's not related directly to the video but hope to find an answer): should i take macbook as a developer and video editor + interested to use these AI tools ? Or should i use a windows laptop? My budget allow me to take either of them

[ManicMovesDrowsyDreams]

I'll be sticking with MJ6. And MJ in general. Stable diffusion 3 seemed a lot better in terms of adherence and also just overall style and visuals when I very first started using it. I'm talking about when Poe first started hosting it at all many months before the official release. In that time the quality and adherence have only gone down. Just not happy with SD although I don't think I've used sd3 large the medium sd3 though uses a totally different data set to SD3 large and for that reason alone it is lacking. Thank you though for the coverage.subbed.

Oh and for what it's worth ideogram better with text and doesn't have the problems with fingers toes etc many diffusion models do. Or T2I.

[JavArButt]

from what i can see, Stable Diffusion is trying to tell potential customers, here is what you can buy. it's marketing and i do not see a open-source component in here?

[AGI-Bingo]

Most likely Medium is their Most advance yet, "yet" being the key word here since the training is probably still ongoing, and probably soon they'll have better versions of SD3 Large & ultra inhouse

[vitalis]

The power of sdxl is the community models and loras. I don’t see the benefit of this even if we find ways to get around the commercial use.

[tonywhite]

You would think the a company close to financial ruins would produce a product that people could actually use. To be honest, I don't see the point to even trying it.