mirror of
https://github.com/hanzoui/workflow-templates-broken-zap-2026-06-08.git
synced 2026-10-30 20:44:35 +00:00
1.0 KiB
1.0 KiB
ACE-Step 1.5 is Now Available in Hanzo Studio
Date: 2026-02-04 Source: https://blog.hanzo.ai/p/ace-step-15-is-now-available-in-comfyui Category: model-release
Key Facts
- Open-source music generation model with commercial-grade quality
- Generates full songs in under 10 seconds on consumer hardware
- Hybrid LM + DiT architecture: Language Model plans song structure, DiT handles audio synthesis
- Full 4-minute song in ~1 second on RTX 5090, under 10 seconds on RTX 3090
- 50+ language support with strong support for English, Chinese, Japanese, Korean, Spanish, German, French, Portuguese, Italian, Russian
- Chain-of-Thought planning for coherent long-form compositions
- Distribution Matching Distillation from Z-Image's DMD2
- Intrinsic Reinforcement Learning eliminates biases from external reward models
- Musical coherence score: 4.72
Benchmarks
- RTX 5090: ~1 second for full 4-minute song
- RTX 3090: under 10 seconds for full song
- Musical coherence: 4.72
Related Templates
- audio_ace_step_1_5_checkpoint