The full DeepSeek-R1 is a 671B-parameter mixture-of-experts model with 37B active parameters; it is far beyond a single consumer PC. The distilled versions (built on Llama and Qwen) run on ordinary hardware.
| Model | Params (total / active) | Est. memory, Q4 @ 8k ctx | Est. memory, Q8 @ 8k ctx | Cheapest fit in our data (Q4) |
|---|---|---|---|---|
| DeepSeek-R1-Distill-Llama-8B | 8.03B / 8.03B | 7.0 GB | 10.6 GB | GeForce RTX 5060 ($299, GPU launch MSRP) |
| DeepSeek-R1-Distill-Qwen-14B | 14.8B / 14.8B | 11.5 GB | 18.2 GB | GeForce RTX 5060 Ti 16GB ($429, GPU launch MSRP) |
| DeepSeek-R1-Distill-Qwen-32B | 32.8B / 32.8B | 22.5 GB | 37.3 GB | GeForce RTX 3090 ($1,499, GPU launch MSRP) |
| DeepSeek-R1-Distill-Llama-70B | 70.6B / 70.6B | 46.8 GB | 78.7 GB | GMKtec EVO-X2 (128GB / 1TB) ($3,499.99, system, listed price) |
| DeepSeek-R1 671B (MoE) | 671B / 37B | 412.6 GB | 715.4 GB | No single device in our dataset |
Parameter counts: Hugging Face: deepseek-ai/DeepSeek-R1 model card. Memory figures are our estimates (see methodology), not measurements.
Model details: Hugging Face: deepseek-ai/DeepSeek-R1 model card.
Check your own hardware
Can My PC Run It?
Loading…