The concept is simple: bring your API key, generate images, and share your results with others. Every image generated through the platform becomes part of a searchable public gallery. This way, we can all learn from each other's experiments and understand what works (and what doesn't).

Here's what you can do:
- Generate images using your SD3.5 API key
- Explore what others have created
- Learn from the community's discoveries
- Share your insights and findings

Remember: this is a public gallery. Everything generated here becomes part of our shared knowledge base.

What is the purpose of this site?

I built this site to help fellow AI enthusiasts figure out what Stability AI's new SD3.5 model can really do. Testing new AI models can be expensive and time-consuming when working alone. At $0.06 per image, running enough experiments to understand SD3.5's true capabilities adds up quickly.

That's why I created this platform – so we can learn together rather than in isolation.

Evaluating new AI models isn't straightforward:
- Marketing examples tend to show only the best results
- Parameters like seeds, steps, and schedulers can significantly impact output to objectively evaluate quality
- Individual experiments often stay hidden in personal projects
- It's hard to know if you're using the model to its full potential

I believe that by sharing our experiments openly, we can build a better understanding of what SD3.5 can and can't do.

Why?

- Your API key is stored only in your browser's local storage, never on any server
- When you generate images, the request needs to go through a server due to CORS restrictions that prevent direct browser-to-API calls (I tried, didn't work)
- Your API key is only used in the moment of making these requests - it isn't stored, logged, or used for anything else
- You can remove your API key from your browser anytime by clearing the cache

About your API keys

This is an independent project, not affiliated with Stability AI. I'm just a developer curious about understanding their latest model better, and I believe in the power of collective learning.

SD3.5

SD3.5 Community Gallery

Stable Diffusion 3 (SD3)

Stable Diffusion 3.5 Large

Stable Diffusion 3.5 Large Turbo

About This Site