Orynos / Audio generation
Start with an idea.
Make it music.
Describe the sound, add your words and create on your own device.
Audio generation is currently available to administrators only. You can explore the models below; downloads and generation remain locked.
Desktop GPU recommended. Model downloads can be several GB. Check the displayed download size before continuing.
Generation stays on this device. Prompts and lyrics are not uploaded for inference. Model files are downloaded separately; browser storage may be cleared.
Preview access is restricted
SessionNo model download or generation is running.
03Your creation
Listen. Save. Make it yours.Your next idea starts here.
Generated audio will appear here when it is ready. Nothing has been generated yet.
AI-generated audio
From an idea to a first listen.
Listen to the full result before using it. Save your audio before closing the page; the browser is not a permanent archive.
Models & provenance
Adapted for Orynos.
The browser runtime, model packaging, verified downloads, memory controls and generation interface are adapted for Orynos. These are existing models, not newly trained or fine-tuned Orynos weights. Community conversions and quantizations retain their original credits and licenses.
Orynos Light uses ACE-Step1.5
ACE-Step1.5 · MIT license. Original model attribution and the relevant notices are retained in the packaged runtime.
Orynos Pro uses MiniMax-Music3
MiniMax-Music3 Community License. Its usage conditions and AI-output disclosure requirements apply.
Commercial use depends on the model terms and your rights to all inputs. Spotify and your distributor make their own acceptance decisions. Use authorized lyrics and other source material; do not impersonate an identifiable artist or voice without permission. No distribution approval is guaranteed.