Blog Post: ComfyUI Adds Stable Audio 3.0 Support for Music and SFX Workflows
Introduction
The ComfyUI Blog announces native support for Stability AI’s Stable Audio 3.0 models, bringing variable-length audio generation—from brief sound effects to full-length musical tracks—into existing ComfyUI workflows. This update is now available for download and offers creators more flexible, commercially licensed audio tools.
What ComfyUI Is Announcing
As of May 20, 2026, ComfyUI includes Stable Audio 3.0 as part of its template library under the Audio category. The integration supports three optimized model variants:
- Small-SFX: For short sound effects and ambient clips (up to 2 minutes)
- Small-Music: On-device-friendly loops and music clips (up to 2 minutes)
- Medium: Longer, more structured musical compositions (up to ~6 minutes)
These models run efficiently on CPUs for smaller tasks and scale up to GPU-accelerated setups for longer outputs.
Key Features in This Release
- Commercially Licensed Data: All models are trained on fully licensed music content.
- Flexible Output Length: From 3-second one-shots to multi-minute musical pieces.
- On-Device Performance: Small models support lightweight CPUs for basic use cases.
- Enhanced Musicality: Medium models deliver richer arrangements over extended durations.
Why This Matters for Workflow Creators
For audio-focused template developers and creators using ComfyUITemplates.com, this update opens new creative possibilities:
- Modular Audio Design: Easily integrate generated SFX, loops, or full tracks into visual projects.
- Custom Prompt Examples: Supports detailed prompts like genre, tempo, instrumentation, and mood.
- Example:
Synthwave 80s retro track with arpeggiated synth leads, analog pads, electric bass...
- Example:
- Template Flexibility: New templates can be built around different model variants depending on desired output length and system specs.
How This Affects ComfyUI Templates and Apps
Template authors can now offer pre-configured Stable Audio 3.0 workflows tailored to specific use cases:
- Short-form Content: Ideal for game assets, UI sounds, or cinematic stings.
- Full-Length Tracks: Perfect for lo-fi beats, synthwave, or ambient background scores.
- Resource-Aware Templates: Provide options for CPU-only vs GPU setups via model selection.
To get started:
- Update to ComfyUI v0.22.0 or use Comfy Cloud.
- Navigate to Template > Audio > Stable Audio 3.0.
- Follow prompts to download models locally if needed.
- Input your prompt and set duration in seconds before running.
Further Reading / Official Resources
This update marks a significant step forward for audio integration within the ComfyUI ecosystem, offering both accessibility and depth for creators exploring generative sound design.