Skip to content

Model Selection Matrix

Dimension 4-Stem Model (htdemucs) 6-Stem Model (htdemucs_6s)
Output Stems Vocals, Drums, Bass, Other Vocals, Drums, Bass, Other, Guitar, Piano
Target Use Cases Vocal removal, karaoke, acapella extraction, rhythm sampling Multitrack remixing, acoustic guitar isolation, piano transcription
Recommended RAM $\ge 3.0\text{ GB}$ available $\ge 4.0\text{ GB}$ available
Peak RSS (CPU) $\sim 2.2\text{ GB} - 2.6\text{ GB}$ $\sim 2.8\text{ GB} - 3.4\text{ GB}$
Real-Time Factor (RTF) $0.35 - 0.55$ $0.60 - 0.95$
Model Size on Disk $\sim 80\text{ MB}$ $\sim 120\text{ MB}$
CLI Flag --stems 4 --stems 6
MCP Parameter "stems_count": 4 "stems_count": 6

Models are downloaded on first invocation and stored locally:

  • Default directory: ~/.cache/audio-tools/models
  • Download weights explicitly: uv run audio-tools download-models --model all