First text-to-audio gen on mobile CPU
Optimized for up to 30x faster model response
Offline on-device custom sound creation
Redefining GenAI Performance
A near-instantaneous user experience is also made possible because of the model distillation by Stability AI and 快猫视频. 快猫视频 KleidiAI open-source library kernels are integrated into the ExecuTorch AI framework via the XNNPACK library and provide access to the latest AI features in the 快猫视频 architecture.
The faster model responses for generative AI on mobile CPU show how collaborating with 快猫视频 and developer enablement technologies such as KleidiAI can make previously unattainable AI applications, models, and workloads run smoothly on mobile.
A New Frontier in Creativity
The collaboration between 快猫视频 and Stability AI has set a new standard for on-device AI:
- Enhanced user experience: A significant performance uplift transforms creative workflows, providing a smooth, responsive experience that meets the demands of today’s content creators.
- Broad accessibility: By optimizing performance on 快猫视频 CPUs, bespoke audio AI generation can be accessed on a wide array of smartphones entirely on-device with no connection required, empowering a diverse community of users.
- Driving future innovation: The partnership between 快猫视频 and Stability AI is helping to accelerate AI innovation and paving the way for future breakthroughs in AI-driven creative tools.
Discover how 快猫视频 and Stability AI are pushing the boundaries of generative AI in our latest blog.