Help Center

Stable Diffusion is an open-source text-to-image model. stable diffusion offers online generation, SDXL / SD 3.5 guides, prompt library, and tutorials for AI art.

stability.ai is the official org. stable diffusion is an independent community focused on Stable Diffusion generation, prompts, ControlNet / LoRA tutorials, and tools — not affiliated with Stability AI.

Yes. The generator accepts multilingual prompts, and our library includes copy-ready Stable Diffusion / SDXL templates plus negative-prompt guidance.

Workflows cover SD 1.5, SDXL, SD 3.5 and Turbo variants, plus ControlNet, LoRA, VAE, inpaint and upscale. Pick model and style on the generate page.

Prompts describe the scene (subject, light, lens, style). Negatives list what to avoid (blur, bad hands, watermarks). Copy examples from our prompt library into the generator.

LoRA injects character/style; ControlNet guides composition via depth, pose, or edges. Tutorials cover stacking weights; the generator supports reference img2img.

Img2img transforms a reference; inpaint fixes regions; outpaint expands the canvas. The tools page links every mode back to the generator.

Many open SD base models allow commercial use; third-party checkpoints/LoRAs may differ. Check licenses and our Terms before commercial projects.

Path: tutorials → copy an SDXL prompt → open AI Generate → iterate with the gallery. Browser-only — no local GPU install required.

SDXL has a mature LoRA ecosystem for style exploration; SD 3.5 follows complex prompts better. Compare on the generate page and read our SD 3.5 vs SDXL post.

Strengthen negatives, reduce empty quality spam, lock a seed while iterating, or inpaint hands. Use the inpaint tutorial and negative templates.

Browsing tutorials, prompts, and models is free. Generation limits follow the live generate page. Learn first with the library and gallery, then generate as needed.