Skip to content
localmodel.run

audio model · orpheus · Linux

Can I run Orpheus 3B on Nvidia DGX Spark (128GB)?

Compatibility verdict memory check
Yes, it runs fast on this GPU

Yes. Orpheus 3B runs on Nvidia DGX Spark (128GB) at Q4_K_M GGUF (~4 GB of ~112 GB usable).

Needs ~4 GB Device usable ~112 GB

Runs at Q4_K_M GGUF using ~4 GB of ~112 GB usable.

Peak memory
~4 GB
Usable on device
~112 GB
Device memory
128 GB
Quant
Q4_K_M GGUF
Share on X Share on Reddit

How to run it

Use llama.cpp or LM Studio at Q4_K_M GGUF. It is light enough to run on CPU; a GPU just makes it faster.

Model orpheus
Type
Text to speech
Parameters
3B
Peak memory
~4 GB at Q4_K_M GGUF
License
Apache-2.0
Full Orpheus 3B requirements →
Device Linux
Memory
128 GB unified
Usable for weights
~112 GB
Power draw
~140 W
Best runtime
llama.cpp (CUDA) / Ollama (CUDA)
Best models for Nvidia DGX Spark (128GB) →

You could also run

Run Orpheus 3B on other hardware

Nvidia GeForce RTX 2060 (6GB)Apple M1 (8GB)8GB RAM Laptop (CPU/iGPU only)Nvidia GeForce RTX 3060 Ti (8GB)Nvidia GeForce GTX 1070 (8GB)Nvidia GeForce RTX 3060 (12GB)

FAQ

Can Nvidia DGX Spark (128GB) run Orpheus 3B?

Yes. Orpheus 3B runs on Nvidia DGX Spark (128GB) at Q4_K_M GGUF (~4 GB of ~112 GB usable).

How much memory does Orpheus 3B need?

Nvidia DGX Spark (128GB) has room to spare. At Q4_K_M GGUF the realistic peak is ~4 GB of memory.

What do I use to run Orpheus 3B locally?

Orpheus 3B runs in llama.cpp or LM Studio (among others). It runs on CPU, so no GPU is required.

Sources

VRAM figures are sourced peak-usage anchors at the noted quant; catalog updated 2026-11-01. See methodology.