Stable Audio 3.0 released! Opening new frontiers in music generation with open weights.
Stable Audio 3.0 Released: A Family of Open-Weight Music Models
Original: Stable Audio 3.0, the model family built with open-weight models<br/> — Stability AI
Importance: 新しいオープンウェイトの音楽生成モデルがリリースされ、多くのユーザーに影響を与えるため
Summary
Stability AI has announced Stable Audio 3.0, a family of open-weight music models trained on fully licensed data. Users can freely distribute and commercialize their outputs, with four models available for various use cases. This release aims to foster creative experimentation within the music community, anticipating innovations in generative audio.
Key Points
- Stable Audio 3.0 is an open-weight music model
- Users can freely commercialize their outputs
- Four models cater to different use cases
- Trained on fully licensed data
- Promotes innovation in the music community
View developer notes (APIs, breaking changes, migration)
Stable Audio 3.0 features four models tailored for different use cases: Stable Audio 3.0 Small SFX for on-device sound effects, Stable Audio 3.0 Small for full music composition, Stable Audio 3.0 Medium for enhanced musicality, and Stable Audio 3.0 Large for low-latency generation. All models are trained on fully licensed data and offer ownership of outputs under the Stability AI Community License. The Enterprise License is available for organizations with annual revenues exceeding $1 million, providing commercial coverage.
Outlet: Stability AI
This article is an AI-generated summary (OpenAI GPT-4o-mini) of publicly available information from Anthropic, OpenAI, Google, Meta, Mistral, DeepSeek, Sakana, and other vendors. The original source URL is always provided in accordance with fair-use citation requirements. Summaries are AI-generated and may contain mistranslations or misinterpretations. Always verify details with the original source.