DeepSeekDeepSeek R1Qwen2ForCausalLM

DeepSeek R1 Distill Qwen 1.5B — Hardware Requirements & GPU Compatibility

ChatReasoningFunctions

DeepSeek R1 Distill Qwen 1.5B is the smallest model in the R1 distillation family, packing chain-of-thought reasoning capabilities into just 1.5 billion parameters using the Qwen 2.5 architecture. It represents an ambitious attempt to bring structured reasoning to the smallest practical model size. At this scale, the model can run on virtually any modern GPU and even on CPU-only setups with acceptable speed. While its reasoning depth is naturally limited compared to its larger siblings, it still demonstrates structured thinking patterns that set it apart from generic models of similar size.

1.8B params · dense131K contextMITJan 2025732.2K downloads1.7K likes180.3K quant downloads

Your hardware

Verdicts below are for the reference RTX 4090. Set yours to personalise the page.

Detection reads your browser's GPU name with no data sent anywhere.

Runs great on a 24 GB GPU

Runs from 8 GB (Q4_K_M · 128K)Best on 24 GB (Q4_K_M · ~461 tok/s · 128K)Macs from 8 GB (Q4_K_M · ~51 tok/s)

MIT

Reference GPU · RTX 4090 24 GB

Estimates, ±20% · how we estimate

Run DeepSeek R1 Distill Qwen 1.5B

One recommended way. Every other method, OS and quant is in the step-by-step guide.Each way opens a step-by-step guide for your OS and quant.

Ways to runOther ways

each opens step-by-step

Run in cloud

Runs on RTX 3060 12GB (10 GB headroom) · Q4_K_M

Generation speed: ~165 tok/s; generation speed
Cost per 1M output tokens: $0.10; per 1M output tokens

Deployon Vast.aiBest valuefrom $0.061/hr

Compare GPUs →

Run DeepSeek R1 Distill Qwen 1.5B

1.8 GB VRAM at 16K context · fits the RTX 4090 (22.2 GB free)

Works on macOS, Windows and Linux.

Install Ollama (Linux)
$curl -fsSL https://ollama.com/install.sh | sh
Install Ollama (macOS / Windows)
Download and run the app; it installs the ollama command.
Download Ollama ↗
Run Q4_K_M
$ollama run deepseek-r1:1.5b
Official Ollama library build. For the exact Q4_K_M file from unsloth/DeepSeek-R1-Distill-Qwen-1.5B-GGUF: ollama run hf.co/unsloth/DeepSeek-R1-Distill-Qwen-1.5B-GGUF:Q4_K_M
Raise the context window
Type it inside the chat. For the API server instead, start it with OLLAMA_CONTEXT_LENGTH=16384 ollama serve.
$/set parameter num_ctx 16384
A larger context needs more memory than the weights alone.
Use the API
Ollama serves its API on http://localhost:11434.

Works on macOS, Windows and Linux.

Install LM Studio
Desktop app for macOS (Apple silicon), Windows and Linux.
lmstudio.ai ↗
Download the model
In the app press Ctrl/Cmd + Shift + M and paste https://huggingface.co/unsloth/DeepSeek-R1-Distill-Qwen-1.5B-GGUF, then pick Q4_K_M. Or use the CLI (run LM Studio once first):
$lms get https://huggingface.co/unsloth/DeepSeek-R1-Distill-Qwen-1.5B-GGUF@Q4_K_M
Open on Hugging Face ↗
Load and chat
Open the model loader, select the model and set Context Length to 16384. Leave GPU offload at maximum if it fits.
Start the local server (optional)
$lms server start

Works on macOS, Windows and Linux.

Install llama.cpp (macOS)
MacPorts, Nix and conda-forge also work.
$brew install llama.cpp
Install options ↗
Install llama.cpp (Windows)
$winget install llama.cpp
Install llama.cpp (Linux)
Download the prebuilt llama-*-bin-ubuntu-* archive for your GPU backend (CPU, Vulkan) from the releases page, unpack it and run the binaries from that folder. Homebrew (brew install llama.cpp) and conda-forge also work on Linux.
llama.cpp releases ↗
Start the server
Downloads the model on first run. The chat UI and an OpenAI-compatible API are at http://localhost:8080.
$llama-server -hf unsloth/DeepSeek-R1-Distill-Qwen-1.5B-GGUF:Q4_K_M -c 16384 -ngl 99 --port 8080

Works on macOS, Windows and Linux.

Option A: Docker Model Runner
Docker Desktop: Settings > AI > Enable Docker Model Runner. Docker Engine on Linux: install the docker-model-plugin package.
$docker model run hf.co/unsloth/DeepSeek-R1-Distill-Qwen-1.5B-GGUF:Q4_K_M
Model Runner docs ↗
The quant tag after the repo is how Hugging Face cards write it; if your Docker version rejects it, drop the tag and pick the quant by repo.
Option B: llama.cpp server image (NVIDIA)
Needs the NVIDIA Container Toolkit. Models are cached in the llama-cache volume. Open http://localhost:8080.
$docker run --gpus all -p 8080:8080 -v llama-cache:/root/.cache ghcr.io/ggml-org/llama.cpp:server-cuda -hf unsloth/DeepSeek-R1-Distill-Qwen-1.5B-GGUF:Q4_K_M -c 16384 -ngl 99 --host 0.0.0.0 --port 8080
Image tags (cuda13, vulkan, rocm) ↗
Cache path inside the image is not in the docs; remove the -v flag if the volume complains.

Linux only.

Install vLLM
Linux only; on Windows run it inside WSL2. NVIDIA GPU shown; AMD ROCm, Intel and CPU builds have their own install pages.
$uv venv --python 3.12 --seed
$source .venv/bin/activate
$uv pip install vllm --torch-backend=auto
vLLM installation ↗
Serve the model
OpenAI-compatible API at http://localhost:8000/v1.
$vllm serve deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B --max-model-len 16384
Runs the original weights, not the Q4_K_M GGUF; needs ~4 GB of GPU memory.

macOS only.

MLX supports only macOS, so follow these steps for that setup.

Install mlx-lm
$pip install mlx-lm
Chat in the terminal
$mlx_lm.chat --model mlx-community/DeepSeek-R1-Distill-Qwen-1.5B-4bit
Downloads the model on first run. The MLX repo has its own quantisation; it may not match the quant selected above.
Or start a local server
Listens on http://localhost:8080.
$mlx_lm.server --model mlx-community/DeepSeek-R1-Distill-Qwen-1.5B-4bit

Works on macOS, Windows and Linux.

Install Jan
jan.ai ↗
Find the model
Open the Hub tab, paste this model ID in Search, and choose Q4_K_M if offered.
$unsloth/DeepSeek-R1-Distill-Qwen-1.5B-GGUF
Use the local API (optional)
Settings > Local API Server > Start Server. It listens on http://127.0.0.1:1337.

Rent GPUs hourly to run Q4_K_M right there. Pricing comes from partner marketplaces and varies.

Available GPU rentals for Q4_K_M, prioritized by value
GPU	Speed	$/1M tokens	Rent
RTX 3060 12GBCheapestCheapest10 GB free	~165 tok/s	$0.10	Deploy on Vast.aifrom $0.061/hr (opens in new tab)	10 GB free · ~165 tok/s · $0.10 / 1M tokens
RTX 3060 Ti (8 GB)6 GB free	~205 tok/s	$0.093	Deploy on Vast.aifrom $0.069/hr (opens in new tab)	8 GB · 6 GB free · ~205 tok/s · $0.093 / 1M tokens
RTX 4060 (8 GB)6 GB free	~125 tok/s	$0.19	Deploy on Vast.aifrom $0.087/hr (opens in new tab)	8 GB · 6 GB free · ~125 tok/s · $0.19 / 1M tokens
RTX 3070 (8 GB)6 GB free	~205 tok/s	$0.12	Deploy on Vast.aifrom $0.087/hr (opens in new tab)	8 GB · 6 GB free · ~205 tok/s · $0.12 / 1M tokens
RTX 3090 (24 GB)22 GB free	~429 tok/s	$0.14	Deploy on RunPodfrom $0.22/hr (opens in new tab)	24 GB · 22 GB free · ~429 tok/s · $0.14 / 1M tokens
RTX 5090 (32 GB)30 GB free	~820 tok/s	$0.23	Deploy on RunPodfrom $0.69/hr (opens in new tab)	32 GB · 30 GB free · ~820 tok/s · $0.23 / 1M tokens
A100 80GB78 GB free	~933 tok/s	$0.41	Deploy on RunPodfrom $1.39/hr (opens in new tab)	78 GB free · ~933 tok/s · $0.41 / 1M tokens
RTX PRO 6000 (96 GB)94 GB free	~731 tok/s	$0.64	Deploy on RunPodfrom $1.69/hr (opens in new tab)	96 GB · 94 GB free · ~731 tok/s · $0.64 / 1M tokens

Compare GPUs →

How Much VRAM Does DeepSeek R1 Distill Qwen 1.5B Need?

Select a quantization level; the listed hardware adheres to it.

VRAM	+32K	+128K	File	Quality	On RTX 4090 (reference)
0.8 GB	1.7 GB	4.6 GB	~0.5 GB	2-bit · i-quant extreme compression — Importance-weighted 2-bit, extreme compression — significant quality loss	Fits: 128K context · ~771 tok/s
0.9 GB	1.8 GB	4.6 GB	0.6 GB	2-bit · i-quant extra small — Importance-weighted 2-bit, extra small	Fits: 128K context · ~736 tok/s
0.9 GB	1.8 GB	4.6 GB	0.7 GB	2-bit · i-quant small — Importance-weighted 2-bit, small	Fits: 128K context · ~720 tok/s
1.0 GB	1.8 GB	4.7 GB	0.8 GB	2-bit · i-quant medium — Importance-weighted 2-bit, medium	Fits: 128K context · ~683 tok/s
1.1 GB	1.9 GB	4.8 GB	0.9 GB	3-bit · i-quant — Importance-weighted 3-bit	Fits: 128K context · ~624 tok/s
1.1 GB	2 GB	4.8 GB	0.7 GB	2-bit · small K-quant — 2-bit small K-quant	Fits: 128K context · ~612 tok/s
1.1 GB	2 GB	4.8 GB	0.8 GB	3-bit · i-quant extra small — Importance-weighted 3-bit, extra small	Fits: 128K context · ~601 tok/s
1.1 GB	2 GB	4.8 GB	0.9 GB	3-bit · i-quant small — Importance-weighted 3-bit, small	Fits: 128K context · ~590 tok/s
1.1 GB	2 GB	4.8 GB	0.8 GB	2-bit · K-quant — 2-bit quantization with K-quant improvements	Fits: 128K context · ~590 tok/s
1.1 GB	2 GB	4.8 GB	0.9 GB	3-bit small — 3-bit small quantization	Fits: 128K context · ~575 tok/s
1.2 GB	2 GB	4.9 GB	0.9 GB	3-bit · i-quant medium — Importance-weighted 3-bit, medium	Fits: 128K context · ~565 tok/s
1.2 GB	2.1 GB	4.9 GB	0.9 GB	3-bit medium — 3-bit medium quantization	Fits: 128K context · ~533 tok/s
1.3 GB	2.1 GB	5 GB	1.1 GB	4-bit · legacy — 4-bit legacy quantization	Fits: 128K context · ~524 tok/s
1.3 GB	2.2 GB	5 GB	1 GB	3-bit large — 3-bit large quantization	Fits: 128K context · ~516 tok/s
1.3 GB	2.2 GB	5 GB	1.1 GB	4-bit · i-quant compact — Importance-weighted 4-bit, compact	Fits: 128K context · ~500 tok/s
1.4 GB	2.2 GB	5.1 GB	1.2 GB	4-bit · legacy+ — 4-bit legacy quantization with offset	Fits: 128K context · ~482 tok/s
1.4 GB	2.2 GB	5.1 GB	1.1 GB	4-bit small — 4-bit small quantization	Fits: 128K context · ~482 tok/s
1.4 GB	2.2 GB	5.1 GB	1.1 GB	4-bit · i-quant non-linear — Importance-weighted 4-bit, non-linear	Fits: 128K context · ~482 tok/s
1.4 GB	2.3 GB	5.1 GB	1.1 GB	4-bit medium — 4-bit medium quantization — most popular sweet spot	Fits: 128K context · ~461 tok/s
1.4 GB	2.3 GB	5.2 GB	1.3 GB	4-bit large — 4-bit large quantization	Fits: 128K context · ~452 tok/s
1.5 GB	2.4 GB	5.2 GB	1.3 GB	5-bit · legacy — 5-bit legacy quantization	Fits: 128K context · ~446 tok/s
1.6 GB	2.5 GB	5.3 GB	1.3 GB	5-bit small — 5-bit small quantization	Fits: 128K context · ~415 tok/s
1.6 GB	2.5 GB	5.3 GB	1.3 GB	5-bit medium — 5-bit medium quantization — good quality/size tradeoff	Fits: 128K context · ~404 tok/s
1.6 GB	2.5 GB	5.4 GB	1.4 GB	5-bit large — 5-bit large quantization	Fits: 128K context · ~397 tok/s
1.8 GB	2.7 GB	5.5 GB	1.5 GB	6-bit, very good — 6-bit quantization, very good quality	Fits: 128K context · ~360 tok/s
2.1 GB	3 GB	5.8 GB	1.9 GB	8-bit, near-lossless — 8-bit quantization, near-lossless	Fits: 128K context · ~306 tok/s
3.9 GB	4.8 GB	7.6 GB	3.6 GB	16-bit · bfloat — Brain floating point 16 — preferred for training	Fits: 128K context · ~168 tok/s

Selected Q4_K_M

4-bit medium quantization

1.8GBat 16K context

Weights: 1.1 GB
KV cache: 0.5 GB
Overhead: 0.3 GB

Fits RTX 4090 (reference) · 22.2 GB free · ~461 tok/s

≥50 tok/s needs ≥109 GB/s · RTX 5090 ~820, RTX 3090 Ti ~461

Which Hardware Can Run DeepSeek R1 Distill Qwen 1.5B?

Q4_K_M · 1.4 GB

DeepSeek R1 Distill Qwen 1.5B (Q4_K_M) requires 1.4 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 2+ GB is recommended. Using the full 131K context window can add up to 3.7 GB, bringing total usage to 5.1 GB. 52 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti. 163 devices with unified memory can run DeepSeek R1 Distill Qwen 1.5B, including Mac Studio (M3 Ultra, 256GB), Mac Studio (M3 Ultra, 512GB).

By VRAM tier · typical GPUs and Macs at Q4_K_M

Typical hardwareQuantVRAMtok/sContext that fits

8 GB · RTX 4060Q4_K_M · 1.8 GB · ~125 tok/s · 128K1.8~125128K8 GB tier →

12 GB · RTX 3060 12GBQ4_K_M · 1.8 GB · ~165 tok/s · 128K1.8~165128K12 GB tier →

16 GB · RTX 5060 Ti 16GBQ4_K_M · 1.8 GB · ~205 tok/s · 128K1.8~205128K16 GB tier →

24 GB · RTX 4090referenceQ4_K_M · 1.8 GB · ~461 tok/s · 128K1.8~461128K24 GB tier →

Mac 32 GB · Mac Mini M4Q4_K_M · 1.8 GB · ~59 tok/s · 128K1.8~59128KMacs →

Mac 64 GB · Mac Studio M4 MaxQ4_K_M · 1.8 GB · ~269 tok/s · 128K1.8~269128KMacs →

How good is DeepSeek R1 Distill Qwen 1.5B?

llmrun Score blends public benchmark boards into one number; 100 = the average open model.

Not enough data yet for an llmrun Score.

A Score needs results on enough independent boards. How the Composite Score is built

Benchmarks

GPQA Diamond
Epoch AI · Jan 2025Independent
33.6%
#80/100 open
Independent
AIME 2024/2025
Epoch AI · Jan 2025Independent
21.4%
#55/88 open
Independent

All 2 DeepSeek R1 Distill Qwen 1.5B benchmarks

Alternatives to DeepSeek R1 Distill Qwen 1.5B

DeepSeek R1 family · 13 models →

Similar models

6tok/s on RTX 4090

Model	Score	VRAM	tok/s	Why it is similar
D1 2BLiquid AI	—	2.3 GB	~423	same size
LFM3 1.7BLiquid AI	—	2.1 GB	~455	same size
Qwen 1 8BAlibaba	—	1.9 GB	~542	same size
JiRackUltra 1BCMSManhattan	—	1.8 GB	~461	same size
Llm Jp 3.1 1.8B Instruct4llm-jp	—	2.2 GB	~360	same size
Qwen3 1.7B BaseAlibaba	—	3.2 GB	~417	same size

Fine-tunes & variants(2)community builds of DeepSeek R1 Distill Qwen 1.5B

Specifications

Publisher: DeepSeek
Family: DeepSeek R1
Parameters: 1.8B
Architecture: Qwen2ForCausalLM
Context Length: 131,072 tokens
Vocabulary Size: 151,936
Release Date: 2025-01-20
License: MIT

Resources

Frequently Asked Questions

How much VRAM does DeepSeek R1 Distill Qwen 1.5B need?

DeepSeek R1 Distill Qwen 1.5B requires 1.4 GB of VRAM at Q4_K_M, or 3.9 GB at BF16. Full 131K context adds up to 3.7 GB (5.1 GB total).

VRAM = Weights + KV Cache + Overhead

Weights = 1.8B × 4.8 bits ÷ 8 = 1.1 GB

KV Cache + Overhead ≈ 0.3 GB (at 2K context + ~0.3 GB framework)

Fit ratings and hardware model lists check this model with room for a 16K-token context, which needs a little more memory.

KV Cache + Overhead ≈ 4 GB (at full 131K context)

VRAM usage by quantization

Q4_K_M

1.4 GB

Q4_K_M + full context

5.1 GB

Learn more about VRAM estimation →

What's the best quantization for DeepSeek R1 Distill Qwen 1.5B?

For DeepSeek R1 Distill Qwen 1.5B, Q4_K_M (1.4 GB) offers the best balance of quality and VRAM usage. Q4_K_L (1.4 GB) provides better quality if you have the VRAM. The smallest option is IQ2_XXS at 0.8 GB.

VRAM requirement by quantization

IQ2_XXS

0.8 GB~53%

IQ3_S

1.1 GB~75%

Q3_K_L

1.3 GB~86%

Q4_K_M ★

1.4 GB~89%

Q5_0

1.5 GB~90%

BF16

3.9 GB~100%

★ Recommended — best balance of quality and VRAM usage.

Learn more about quantization →

Can I run DeepSeek R1 Distill Qwen 1.5B on a Mac?

Yes — MacBook Air 13" M3 (8 GB) and 145 other Macs can run DeepSeek R1 Distill Qwen 1.5B. Apple Silicon uses unified memory, so the model shares RAM with the system. At Q4_K_M you need at least 1.4 GB of usable unified memory (RAM minus macOS overhead).

Can I run DeepSeek R1 Distill Qwen 1.5B locally?

Yes — DeepSeek R1 Distill Qwen 1.5B can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.

How fast is DeepSeek R1 Distill Qwen 1.5B?

At Q4_K_M, DeepSeek R1 Distill Qwen 1.5B can reach ~3380 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~461 tok/s. Speed depends mainly on GPU memory bandwidth. Real-world results typically within ±20%.

tok/s = (bandwidth GB/s ÷ model GB) × efficiency

Example: NVIDIA B200 → 8000 ÷ 1.42 × 0.65 = ~3662 tok/s

Estimated speed at Q4_K_M (1.4 GB)

NVIDIA B200

~3662 tok/s

NVIDIA GeForce RTX 4090

~461 tok/s

NVIDIA B300

~3662 tok/s

AMD Instinct MI350X

~3380 tok/s

Real-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.

Learn more about tok/s estimation →

What's the download size of DeepSeek R1 Distill Qwen 1.5B?

At Q4_K_M, the download is about 1.07 GB. The full-precision BF16 version is 3.55 GB. The smallest option (IQ2_XXS) is 0.49 GB.

Which GPUs can run DeepSeek R1 Distill Qwen 1.5B?

52 consumer GPUs can run DeepSeek R1 Distill Qwen 1.5B at Q4_K_M. Top options include AMD Radeon RX 6700 XT, AMD Radeon RX 6800, AMD Radeon RX 6800 XT. 52 GPUs have plenty of headroom for comfortable inference.

Which devices can run DeepSeek R1 Distill Qwen 1.5B?

163 devices with unified memory can run DeepSeek R1 Distill Qwen 1.5B at Q4_K_M (1.4 GB), including AMD Ryzen AI 9 HX 370 (Strix Point) Laptop, ASUS Ascent GX10, Beelink GTR9 Pro (Ryzen AI Max+ 395, 128 GB), Framework Desktop (Ryzen AI Max+ 395, 128 GB). Apple Silicon Macs use unified memory shared between CPU and GPU, making them well-suited for local LLM inference.