DeepSeek · 8B · Desktop / import

DeepSeek R1 Distill 8B on a phone

Reasoning-focused — walks through hard problems before answering.

Parameters
8B
Typical Q4
~4.7 GB (Q4)
RAM to budget
~7.2 GB
Engines
GGUF

Get it running

R1 distill on Llama 8B; longer CoT traces. Hugging Face: deepseek-ai/DeepSeek-R1-Distill-Llama-8B.

Load in LM Studio or Ollama, then chat from the phone.

Get LM Mini and load a model that actually fits — on-device, or via Connect to a PC.

Phones that can run it

Tight (close other apps)