DeepSeek · Reasoning Models

DeepSeek-R1 Distill models

Name: DeepSeek-R1 Distill models
Author: DeepSeek

Smaller reasoning models distilled from DeepSeek-R1 into Qwen and Llama bases, from 1.5B to 70B, bringing chain-of-thought reasoning to local hardware.

Official page

Run it locally

With Ollama installed, download and start the default version with:

ollama run deepseek-r1:8b

Variant	Download
deepseek-r1:8b	5.2GB
deepseek-r1:1.5b	1.1GB
deepseek-r1:7b	4.7GB
deepseek-r1:14b	9.0GB
deepseek-r1:32b	20GB
deepseek-r1:70b	43GB

Memory: About 5.2GB for the default deepseek-r1:8b version (4-bit)
Formats: Safetensors (Hugging Face), GGUF (Ollama, llama.cpp, LM Studio)

The download size is roughly the memory (RAM or GPU VRAM) needed to load a variant; allow extra for long contexts. See our hardware guide.

Overview

Alongside DeepSeek-R1, DeepSeek released six distilled models fine-tuned on reasoning data generated by R1, built on Qwen2.5 and Llama 3 bases.

Highlights

Sizes from 1.5B to 70B
Step-by-step reasoning distilled from R1
Based on Qwen2.5 and Llama 3
Popular for local reasoning

Typical uses

Local math and logic assistants and reasoning research.

Things to consider

Less capable than the full R1 model.
Licenses of the Qwen and Llama base models also apply.
Download sizes refer to the default 4-bit quantized versions in the Ollama library; full-precision weights are larger.

Specifications are taken from the model card on Hugging Face, the Ollama library and the provider’s announcement. Benchmark results depend on evaluation settings; see how AI benchmarks work.

#Reasoning Models #Open Weights #Local AI #DeepSeek