Rigsmith

Build for… /

Build for AI

Fine-tuning language or image models (LoRA/QLoRA), and training your own models in PyTorch.

Budget

Goal: Fine-tune ~8B models

$2,108
  • QLoRA fine-tunes of models up to about 8 billion parameters
  • Trains about 1.7× as fast as a typical entry gaming PC (Ryzen 5 7600 + RTX 5060)
  • 64 GB of memory: plenty for datasets and model loading
Processor
6C/12T
Graphics
16 GB card
Memory
64 GB DDR5
Storage
2 TB
Peak power
~320 W

Mid-range

Best value

Goal: Faster fine-tuning

$3,587
  • QLoRA fine-tunes of models up to about 8 billion parameters
  • Trains about 2.3× as fast as a typical entry gaming PC (Ryzen 5 7600 + RTX 5060)
  • 64 GB of memory: plenty for datasets and model loading
Processor
14C/20T
Graphics
16 GB card
Memory
64 GB DDR5
Storage
2 TB
Peak power
~593 W

High-end

Goal: Fine-tune ~20–30B models

$6,302
  • QLoRA fine-tunes of models up to about 20–30 billion parameters
  • Trains about 3.9× as fast as a typical entry gaming PC (Ryzen 5 7600 + RTX 5060)
  • 64 GB of memory: plenty for datasets and model loading
Processor
12C/24T
Graphics
32 GB card
Memory
64 GB DDR5
Storage
2 TB
Peak power
~789 W

Second-hand value pick

Used card

A second-hand RTX 3090 (24 GB) is the classic budget training card.

$3,284
  • QLoRA fine-tunes of models up to about 14 billion parameters
  • Trains about 2.3× as fast as a typical entry gaming PC (Ryzen 5 7600 + RTX 5060)
  • 64 GB of memory: plenty for datasets and model loading
Processor
14C/20T
Graphics
24 GB card
Memory
64 GB DDR5
Storage
2 TB
Peak power
~583 W

Buying second-hand? Check the card works under load before paying, and use the used PC checker for whole PCs.

What matters

  • Video memory: it sets the size of model you can train at all.
  • An NVIDIA card: PyTorch and most training tools are built around CUDA.
  • 64 GB of system memory and a fast drive for datasets.

What to skip

  • Gaming extras like RGB and high-refresh monitors.

Good to know

  • For big jobs, renting a cloud GPU by the hour can beat buying; a local card is best for frequent, smaller runs.
  • Linux (or WSL2 on Windows) is the smoothest place for training tools.
BUDGET_FINDERPro

The best build for your budget

For Training and fine-tuning: set exactly what you want to spend and we find the complete, compatible build that suits it best at that price.

Sign in and upgrade to use the budget finder

Each build is the cheapest complete, compatible PC from current parts that meets its goal, priced at the best Australian store prices as of 12/11/2026. Speed comparisons are Rigsmith estimates, not benchmarks. Open a build in the builder to swap parts or buy second-hand.