Stability AI Unveils Stable Audio 3 for Subsecond Long Audio Generation
Stability AI, a renowned artificial intelligence company, has officially released its latest generation audio large model Stable Audio3 and simultaneously open-sourced part of the model weights. Built as a latent diffusion model specifically designed for audio generation and editing, this system not only supports high-quality stereo output but also achieves a significant breakthrough in generation speed.
The newly released model family covers a wide range of specifications, from small to large, catering to diverse needs such as music creation and sound effect production. Notably, the model supports variable-length audio generation and introduces an audio editing feature based on internal image completion technology, offering creators unprecedented flexibility.

Innovative Architecture Overcomes Hardware Constraints
Stable Audio3 consists of two core components: a semantic acoustic autoencoder called SAME, and an efficient diffusion transformer. The SAME autoencoder achieves an audio compression ratio of up to 4096x, a breakthrough design that significantly shortens the length of the latent sequence.
Thanks to this efficient compression mechanism, even on ordinary consumer-grade hardware, the model can smoothly run long-duration, large-scale audio generation tasks. This not only significantly lowers the technical barrier for high-quality audio creation but also makes professional-level audio and video production at home possible for individual creators.

Ultra Efficiency Enables Instant Rendering
With support for variable-length technology, the new model's computational cost scales dynamically with the user's required audio duration, completely eliminating the computing power waste caused by fixed-length approaches in the past. In tests on high-performance hardware, the model can render a 20-second audio in about 0.62 seconds and generate a 380-second piece of music in just 1.31 seconds.
Additionally, through an innovative three-stage training process, Stable Audio3 no longer relies on traditional classifier-free guidance technology during inference, thus achieving an ultra-fast single-step forward propagation experience. Currently, the small and medium model weights are publicly available on the Hugging Face platform, while the larger version with stronger performance will be provided through commercial licensing.
Related article
Ctrip Business Travel Unveils AI Ecosystem with 7 Industry Agents and 17 Core Capabilities
Ctrip Business Travel has officially unveiled the "Ctrip Business Travel AI Ecosystem," leveraging large language models to fundamentally restructure corporate travel management. This comprehensive ecosystem addresses four core scenarios: travel arra
Stripe to Acquire OpenRouter as AI Model Routing Gains Momentum
Stripe has agreed to acquire OpenRouter, an AI model-routing platform that gives developers access to hundreds of models through a single interface. The deal adds model selection and routing to Stripe’s existing work around AI usage and token-based b
Former OpenAI Researcher Barrett Zoph Returns to Google to Support Gemini Development
Barret Zoph, who previously abruptly left and joined OpenAI due to a disagreement with Thinking Machines Lab CEO Miriam Mularati, has once again changed employers. Alphabet's Google announced that Zoph will return as Vice President of Research, bring
Related Special Topic Recommendations
Comments (0)
0/500
Stability AI, a renowned artificial intelligence company, has officially released its latest generation audio large model
The newly released model family covers a wide range of specifications, from small to large, catering to diverse needs such as music creation and sound effect production. Notably, the model supports variable-length audio generation and introduces an audio editing feature based on internal image completion technology, offering creators unprecedented flexibility.

Innovative Architecture Overcomes Hardware Constraints
Thanks to this efficient compression mechanism, even on ordinary consumer-grade hardware, the model can smoothly run long-duration, large-scale audio generation tasks. This not only significantly lowers the technical barrier for high-quality audio creation but also makes professional-level audio and video production at home possible for individual creators.

Ultra Efficiency Enables Instant Rendering
With support for variable-length technology, the new model's computational cost scales dynamically with the user's required audio duration, completely eliminating the computing power waste caused by fixed-length approaches in the past. In tests on high-performance hardware, the model can render a 20-second audio in about 0.62 seconds and generate a 380-second piece of music in just 1.31 seconds.
Additionally, through an innovative three-stage training process,
Ctrip Business Travel Unveils AI Ecosystem with 7 Industry Agents and 17 Core Capabilities
Ctrip Business Travel has officially unveiled the "Ctrip Business Travel AI Ecosystem," leveraging large language models to fundamentally restructure corporate travel management. This comprehensive ecosystem addresses four core scenarios: travel arra
Stripe to Acquire OpenRouter as AI Model Routing Gains Momentum
Stripe has agreed to acquire OpenRouter, an AI model-routing platform that gives developers access to hundreds of models through a single interface. The deal adds model selection and routing to Stripe’s existing work around AI usage and token-based b
Former OpenAI Researcher Barrett Zoph Returns to Google to Support Gemini Development
Barret Zoph, who previously abruptly left and joined OpenAI due to a disagreement with Thinking Machines Lab CEO Miriam Mularati, has once again changed employers. Alphabet's Google announced that Zoph will return as Vice President of Research, bring





Home






