audio model · musicgen · Windows
Can I run MusicGen small on Nvidia GeForce RTX 4070 (12GB)?
Yes. MusicGen small runs on Nvidia GeForce RTX 4070 (12GB) at fp32 (~3 GB of ~11 GB usable).
Runs at fp32 using ~3 GB of ~11 GB usable.
- Peak memory
- ~3 GB
- Usable on device
- ~11 GB
- Device memory
- 12 GB
- Quant
- fp32
How to run it
Use AudioCraft or HF Transformers at fp32. It loads as a single model and runs on a GPU.
- Type
- Music generation
- Parameters
- 300M
- Peak memory
- ~3 GB at fp32
- License
- CC-BY-NC-4.0
- Memory
- 12 GB vram
- Usable for weights
- ~11 GB
- Best runtime
- Ollama (CUDA) / vLLM (Linux)
You could also run
Run MusicGen small on other hardware
FAQ
Can Nvidia GeForce RTX 4070 (12GB) run MusicGen small?
Yes. MusicGen small runs on Nvidia GeForce RTX 4070 (12GB) at fp32 (~3 GB of ~11 GB usable).
How much memory does MusicGen small need?
Nvidia GeForce RTX 4070 (12GB) has room to spare. At fp32 the realistic peak is ~3 GB of memory.
What do I use to run MusicGen small locally?
MusicGen small runs in AudioCraft or HF Transformers. It needs a GPU.
Sources
VRAM figures are sourced peak-usage anchors at the noted quant, validated 2026-06-15. See methodology.