Currently, SDXL Turbo is available under a non-commercial research license, limiting its use to personal, non-commercial purposes. Despite facing internal management issues, Stability AI continues to release new products, including Stable Video Diffusion, which can turn still images into short video clips. The company offers a beta demonstration of SDXL Turbo's capabilities on its image-editing platform, Clipdrop, and an unofficial live demo on Hugging Face.
Key takeaways:
- Stability AI has launched Stable Diffusion XL Turbo, an AI image-synthesis model that can rapidly generate imagery based on a written prompt. The model can produce image outputs in a single step, a significant reduction from the 20–50 steps required by its predecessor.
- The model's efficiency is attributed to a technique called Adversarial Diffusion Distillation (ADD) which uses score distillation and adversarial loss to improve the realism of the output.
- SDXL Turbo is currently available under a non-commercial research license, limiting its use to personal, non-commercial purposes. However, Stability AI has expressed openness to commercial applications.
- Despite internal management issues and potential company sale, Stability AI continues to release new products, including Stable Video Diffusion, which can turn still images into short video clips.