Allen AIOLMoOlmo2ForCausalLM

OLMo 2 0325 32B Instruct — Hardware Requirements & GPU Compatibility

Chat

OLMo 2 0325 32B Instruct is a 32.2B-parameter open language model from Allen AI in the OLMo family. It supports a context window of up to 4,096 tokens. At Q4_K_M it needs about 20.18 GB of VRAM — see which GPUs and Macs can run it below.

32.2B params · dense4K contextApache 2.0Mar 202569.3K downloads147 likes1.3K 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.

B

Decent on a 24 GB GPU

Runs from 16 GB (Q3_K_S · 4K)Best on 24 GB (Q4_K_M · ~33 tok/s · 4K)Macs from 24 GB (Q4_K_M · ~11 tok/s)

Apache 2.0

Reference GPU · RTX 4090 24 GB

Estimates, ±20% · how we estimate

Run OLMo 2 0325 32B Instruct

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.

Recommended

Ollama

one command · needs Ollama installed

$ollama run hf.co/unsloth/OLMo-2-0325-32B-Instruct-GGUF:Q4_K_M

Q4_K_M fits fully on the RTX 4090 · 3.3 GB free · 4K context

Download: OLMo-2-0325-32B-Instruct-Q4_K_M.gguf19.48 GBunsloth

install, OS, other quants

Ways to runOther ways

each opens step-by-step

Run in cloud

Runs on RTX 5090 (32 GB) (11 GB headroom) · Q4_K_M

Generation speed: ~58 tok/s; generation speed
Cost per 1M output tokens: $2.39; per 1M output tokens

Deployon Vast.aiBest valuefrom $0.50/hr

Compare GPUs →

or

How Much VRAM Does OLMo 2 0325 32B Instruct Need?

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

Quantization	VRAM	+32K	+128K	File	Quality	On RTX 4090 (reference)
	14.5 GB	—	—	12 GB	2-bit · K-quant — 2-bit quantization with K-quant improvements	Fits: 4K context · ~45 tok/s
	14.9 GB	—	—	14.1 GB	3-bit small — 3-bit small quantization	Fits: 4K context · ~44 tok/s
	16.6 GB	—	—	15.6 GB	3-bit medium — 3-bit medium quantization	Fits: 4K context · ~40 tok/s
	16.9 GB	—	—	18.3 GB	4-bit · legacy — 4-bit legacy quantization	Fits: 4K context · ~39 tok/s
	17.4 GB	—	—	16.9 GB	3-bit large — 3-bit large quantization	Fits: 4K context · ~38 tok/s
	18.2 GB	—	—	17.5 GB	4-bit · i-quant compact — Importance-weighted 4-bit, compact	Fits: 4K context · ~36 tok/s
	19.0 GB	—	—	18.4 GB	4-bit small — 4-bit small quantization	Fits: 4K context · ~35 tok/s
	20.2 GB	—	—	19.5 GB	4-bit medium — 4-bit medium quantization — most popular sweet spot	Fits: 4K context · ~33 tok/s
	21.0 GB	—	—	~20.2 GB	5-bit · legacy — 5-bit legacy quantization	Fits: 4K context · ~31 tok/s
	23 GB	—	—	22.2 GB	5-bit small — 5-bit small quantization	Fits: 4K context · ~29 tok/s
	23.8 GB	—	—	22.9 GB	5-bit medium — 5-bit medium quantization — good quality/size tradeoff	Fits: 2K context · ~28 tok/s
	27.4 GB	—	—	26.4 GB	6-bit, very good — 6-bit quantization, very good quality	Does not fit: needs 27.4 GB
	33.1 GB	—	—	34.3 GB	8-bit, near-lossless — 8-bit quantization, near-lossless	Does not fit: needs 33.1 GB
	65.3 GB	—	—	64.5 GB	16-bit · bfloat — Brain floating point 16 — preferred for training	Does not fit: needs 65.3 GB

Selected Q4_K_M

4-bit medium quantization

20.7GBat 4K context

Weights: 19.3 GB
KV cache: 1.1 GB
Overhead: 0.3 GB

Fits RTX 4090 (reference) · 3.3 GB free · ~33 tok/s

≥50 tok/s needs ≥1.55 TB/s · RTX 5090 ~58

Which Hardware Can Run OLMo 2 0325 32B Instruct?

Q4_K_M · 20.2 GB

OLMo 2 0325 32B Instruct (Q4_K_M) requires 20.2 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 27+ GB is recommended. Using the full 4K context window can add up to 0.5 GB, bringing total usage to 20.7 GB. 7 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti. 122 devices with unified memory can run OLMo 2 0325 32B Instruct, including Mac Studio (M3 Ultra, 256GB), Mac Studio (M3 Ultra, 512GB), Mac Studio M1 Max (32 GB).

By VRAM tier · typical GPUs and Macs at Q4_K_M

Typical hardwareQuantVRAMtok/sContext that fits

8–12 GB · RTX 4060, RTX 3060 12GBDoesn't fit — smallest quant needs 15.1 GB8 GB tier →12 GB tier →

16 GB · RTX 5060 Ti 16GBQ3_K_S · 15.5 GB · ~20 tok/s · 4K15.5~204K16 GB tier →

24 GB · RTX 4090referenceQ4_K_M · 20.7 GB · ~33 tok/s · 4K20.7~334K24 GB tier →

Mac 32 GB · Mac Mini M4Q4_K_M · 20.7 GB · ~4 tok/s · 4K20.7~44KMacs →

Mac 64 GB · Mac Studio M4 MaxQ4_K_M · 20.7 GB · ~19 tok/s · 4K20.7~194KMacs →

How good is OLMo 2 0325 32B Instruct?

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

LMArena Text
LMArena · Oct 2026Arena
1251.7
#130/188 open
Arena

OLMo 2 0325 32B Instruct benchmark results

Alternatives to OLMo 2 0325 32B Instruct

OLMo family · 20 models →

Similar models

6tok/s on RTX 4090

Model	Score	VRAM	tok/s	Why it is similar
Nanbeige4.3 33BNanbeige	—	23.7 GB	~32	same size
Hy Memory 30B A3BTencent	—	19.1 GB	~184	same size
Granite Switch 4.1 30BIBM	—	22.7 GB	~35	same size
Llm Jp 4.1 32B A3b Thinkingllm-jp	—	20.1 GB	~188	same size
Llm Jp 4.1 33B Thinkingllm-jp	—	22.5 GB	~35	same size
Agnes 3.0 FlashAgnes-AI	—	22.1 GB	~35	same size

Specifications

Publisher: Allen AI
Family: OLMo
Parameters: 32.2B
Architecture: Olmo2ForCausalLM
Context Length: 4,096 tokens
Vocabulary Size: 100,352
Release Date: 2025-03-12
License: Apache 2.0

Resources

Frequently Asked Questions

How much VRAM does OLMo 2 0325 32B Instruct need?

OLMo 2 0325 32B Instruct requires 20.2 GB of VRAM at Q4_K_M, or 65.3 GB at BF16. Full 4K context adds up to 0.5 GB (20.7 GB total).

VRAM = Weights + KV Cache + Overhead

Weights = 32.2B × 4.8 bits ÷ 8 = 19.3 GB

KV Cache + Overhead ≈ 0.9 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 ≈ 1.4 GB (at full 4K context)

VRAM usage by quantization

Q4_K_M

20.2 GB

Q4_K_M + full context

20.7 GB

Learn more about VRAM estimation →

Can NVIDIA GeForce RTX 4090 run OLMo 2 0325 32B Instruct?

Yes, at Q5_K_M (23.8 GB) or lower. Higher quantizations like Q6_K (27.4 GB) exceed the NVIDIA GeForce RTX 4090's 24 GB.

What's the best quantization for OLMo 2 0325 32B Instruct?

For OLMo 2 0325 32B Instruct, Q4_K_M (20.2 GB) offers the best balance of quality and VRAM usage. Q5_0 (21.0 GB) provides better quality if you have the VRAM. The smallest option is Q2_K at 14.5 GB.

VRAM requirement by quantization

Q2_K

14.5 GB~75%

Q4_0

16.9 GB~85%

Q4_K_M ★

20.2 GB~89%

Q5_0

21.0 GB~90%

Q5_K_M

23.8 GB~92%

BF16

65.3 GB~100%

★ Recommended — best balance of quality and VRAM usage.

Learn more about quantization →

Can I run OLMo 2 0325 32B Instruct on a Mac?

Yes — Mac Mini M5 Pro (24 GB) and 106 other Macs can run OLMo 2 0325 32B Instruct. Apple Silicon uses unified memory, so the model shares RAM with the system. At Q4_K_M you need at least 20.2 GB of usable unified memory (RAM minus macOS overhead).

Can I run OLMo 2 0325 32B Instruct locally?

Yes — OLMo 2 0325 32B Instruct can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.

How fast is OLMo 2 0325 32B Instruct?

At Q4_K_M, OLMo 2 0325 32B Instruct can reach ~238 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~33 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 ÷ 20.18 × 0.65 = ~258 tok/s

Estimated speed at Q4_K_M (20.2 GB)

NVIDIA B200

~258 tok/s

NVIDIA GeForce RTX 4090

~33 tok/s

NVIDIA B300

~258 tok/s

AMD Instinct MI350X

~238 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 OLMo 2 0325 32B Instruct?

At Q4_K_M, the download is about 19.34 GB. The full-precision BF16 version is 64.47 GB. The smallest option (Q2_K) is 13.70 GB.

Which GPUs can run OLMo 2 0325 32B Instruct?

7 consumer GPUs can run OLMo 2 0325 32B Instruct at Q4_K_M. Top options include NVIDIA GeForce RTX 5090, AMD Radeon RX 7900 XTX, NVIDIA GeForce RTX 3090. 1 GPU has plenty of headroom for comfortable inference.

Which devices can run OLMo 2 0325 32B Instruct?

122 devices with unified memory can run OLMo 2 0325 32B Instruct at Q4_K_M (20.2 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.

Run OLMo 2 0325 32B Instruct

20.7 GB VRAM at 4K context · fits the RTX 4090 (3.3 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 hf.co/unsloth/OLMo-2-0325-32B-Instruct-GGUF:Q4_K_M
Raise the context window
Type it inside the chat. For the API server instead, start it with OLLAMA_CONTEXT_LENGTH=4096 ollama serve.
$/set parameter num_ctx 4096
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/OLMo-2-0325-32B-Instruct-GGUF, then pick Q4_K_M. Or use the CLI (run LM Studio once first):
$lms get https://huggingface.co/unsloth/OLMo-2-0325-32B-Instruct-GGUF@Q4_K_M
Open on Hugging Face ↗
Load and chat
Open the model loader, select the model and set Context Length to 4096. 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/OLMo-2-0325-32B-Instruct-GGUF:Q4_K_M -c 4096 -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/OLMo-2-0325-32B-Instruct-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/OLMo-2-0325-32B-Instruct-GGUF:Q4_K_M -c 4096 -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 allenai/OLMo-2-0325-32B-Instruct --max-model-len 4096
Runs the original weights, not the Q4_K_M GGUF; needs ~71 GB of GPU memory.

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/OLMo-2-0325-32B-Instruct-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 5090 (32 GB)CheapestCheapest11 GB free	~58 tok/s	$2.39	Deploy on Vast.aifrom $0.50/hr (opens in new tab)	32 GB · 11 GB free · ~58 tok/s · $2.39 / 1M tokens
RTX PRO 4500 (32 GB)11 GB free	~29 tok/s	$3.27	Deploy on RunPodfrom $0.34/hr (opens in new tab)	32 GB · 11 GB free · ~29 tok/s · $3.27 / 1M tokens
A40 (48 GB)27 GB free	~22 tok/s	$4.34	Deploy on RunPodfrom $0.35/hr (opens in new tab)	48 GB · 27 GB free · ~22 tok/s · $4.34 / 1M tokens
RTX PRO 4500 (32 GB)11 GB free	~29 tok/s	$3.42	Deploy on Vast.aifrom $0.36/hr (opens in new tab)	32 GB · 11 GB free · ~29 tok/s · $3.42 / 1M tokens
RTX A6000 (48 GB)27 GB free	~25 tok/s	$4.55	Deploy on Vast.aifrom $0.40/hr (opens in new tab)	48 GB · 27 GB free · ~25 tok/s · $4.55 / 1M tokens
A100 40GB19 GB free	~50 tok/s	$3.38	Deploy on Vast.aifrom $0.61/hr (opens in new tab)	19 GB free · ~50 tok/s · $3.38 / 1M tokens
RTX 5090 (32 GB)11 GB free	~58 tok/s	$3.32	Deploy on RunPodfrom $0.69/hr (opens in new tab)	32 GB · 11 GB free · ~58 tok/s · $3.32 / 1M tokens
RTX PRO 4500 Server (32 GB)11 GB free	~26 tok/s	$7.75	Deploy on RunPodfrom $0.72/hr (opens in new tab)	32 GB · 11 GB free · ~26 tok/s · $7.75 / 1M tokens

Compare GPUs →