audio model · bark · Windows
Can I run Bark on Nvidia GeForce RTX 2060 (6GB)?
Yes. Bark runs on Nvidia GeForce RTX 2060 (6GB) at fp32 (~5 GB of ~5 GB usable).
Fits at fp32 (~5 GB of ~5 GB usable) but with little headroom. Close other apps to free a little VRAM before generating.
- Peak memory
- ~5 GB
- Usable on device
- ~5 GB
- Device memory
- 6 GB
- Quant
- fp32
How to run it
Use HF Transformers or suno-ai/bark at fp32. It is light enough to run on CPU; a GPU just makes it faster.
- Type
- Text to speech (and sound effects)
- Parameters
- 900M
- Peak memory
- ~5 GB at fp32
- License
- MIT
- Memory
- 6 GB vram
- Usable for weights
- ~5 GB
- Power draw
- ~160 W
- Best runtime
- Ollama (CUDA) / llama.cpp CUDA
You could also run
Run Bark on other hardware
FAQ
Can Nvidia GeForce RTX 2060 (6GB) run Bark?
Yes. Bark runs on Nvidia GeForce RTX 2060 (6GB) at fp32 (~5 GB of ~5 GB usable).
How much memory does Bark need?
It is a tight fit on Nvidia GeForce RTX 2060 (6GB). At fp32 the realistic peak is ~5 GB of memory.
What do I use to run Bark locally?
Bark runs in HF Transformers or suno-ai/bark (among others). It runs on CPU, so no GPU is required.
Sources
VRAM figures are sourced peak-usage anchors at the noted quant, validated 2026-09-14. See methodology.