Mid-range build · mini-ITX · New parts
Arctic Llama
Budget single-GPU AI workstation
Built by Llama Builds community · Updated
- Total cost
- $1,563
- VRAM
- 12GB
- GPUs
- 1
- System RAM
- 32GB
What can this build run?
With 12GB of VRAM across one GPU, this build comfortably runs models up to Phi-4 (Q3_K_M, ~52 tokens/sec). Our top all-round pick for it is Qwen3.5 9B at Q6_K (~46 tokens/sec).
Parts list
Every part in the Arctic Llama
We may earn a commission when you buy through links on this page. It never changes what we recommend. How we make money
| Part | Price | Buy |
|---|---|---|
GPU NVIDIA GeForce RTX 4070 | $599 | Amazon |
CPU AMD Ryzen 7 7800X3D 4.2 GHz 8-Core | $375 | Amazon |
CPU Cooler Thermalright Peerless Assassin 120 SE White ARGB 66.17 CFM | $37 | Amazon |
Motherboard Gigabyte A620I AX Mini ITX AM5 | $150 | Amazon |
Memory Corsair Vengeance RGB 32GB (2 x 16GB) DDR5-5200 CL40 | $97 | Amazon |
Storage Acer Predator GM7000 2TB M.2-2280 PCIe 4.0 x4 NVMe SSD | $130 | Amazon |
Case Cooler Master MasterBox NR200P Mini ITX Desktop | $100 | Amazon |
Power Supply Lian Li SP850 850W 80+ Gold Fully Modular SFX | $134 | Amazon |
| Total | $1,563 |
Prices at time of build (2025); check current prices. Individual parts add up to $1,622.
Capability
Models this build runs well
Estimated for a single user with 8K context. Best quantization that fits, ranked by our capability score.
Best overall · also coding, reasoning, vision, fastest
Qwen3.5 9BQ6_K · ~46 tokens/sec
Largest model that fits
Phi-4Q3_K_M · ~52 tokens/sec
| Model | Quant | Speed | Fit |
|---|---|---|---|
| Qwen3.5 9B | Q6_K | ~46 tok/s | Runs well |
| Gemma 4 E4B | Q8_0 | ~41 tok/s | Runs well |
| Phi-4 | Q3_K_M | ~52 tok/s | Runs well |
The build
About this build
A white mini-ITX build with an RTX 4070 and a Ryzen 7 7800X3D, made to handle CAD, gaming, programming, and small language models like Gemma 3 from one compact desktop.
Why this build
My first PC build after graduating college! I built this white RGB mini-ITX system to replace my old college laptop. It’s designed for SolidWorks CAD work, gaming, programming, and running small language models like Gemma 3.
The compact form factor fits perfectly on my desk while delivering excellent performance for both professional work and AI experimentation.
Notes and caveats
- This build was previously listed under an RTX 4070 Ti name, but it uses the standard RTX 4070 with 12GB of VRAM.
Specifications
Full specs
- GPU
- RTX 4070
- CPU
- AMD Ryzen 7 7800X3D (8-core, 4.2 GHz)
- Motherboard
- Gigabyte A620I AX (Mini-ITX, AM5)
- Memory
- 32GB (2 x 16GB) DDR5-5200 CL40 (Corsair Vengeance RGB)
- Storage
- Acer Predator GM7000 2TB PCIe 4.0 NVMe SSD
- Power supply
- Lian Li SP850 850W 80+ Gold SFX
- Case
- Cooler Master MasterBox NR200P (Mini-ITX)
- Cooling
- Thermalright Peerless Assassin 120 SE White ARGB
- Form factor
- mini-ITX
- GPU power
- 200W rated
Keep browsing
