Check my GPU

on

Can the NVIDIA GeForce RTX 3090 run Olmo 3.1 32B Think?

85 ptsExcellent (S) · 85 of 100 points: Model quality 32.5/45, Memory fit 24.0/25, Speed 14.2/15, Memory use 15.0/15

Yes, with little memory to spare

Tight fit. At Q4_K_M it needs about 22.2 GB of 24.0 GB usable, with an estimated speed of 35.6 tok/s.

Best quant
Q4_K_M
Generation speed
35.6 tok/s
Prompt processing
795 tok/s
First token (1K prompt)
1.3 s
Largest safe context
27K

Download Olmo 3.1 32B Think

Recommended for the NVIDIA GeForce RTX 3090: Q4_K_M. Every other quantization has its own download link in the table below.

Quickest · Ollama

One command downloads a ready-to-run version and starts a chat.

ollama run olmo-3.1:32b-think-q4_K_M

Get Ollama at ollama.com/download (opens in a new tab) · model page (opens in a new tab)

Quantized files · GGUF

Works in LM Studio, llama.cpp, Jan and Ollama. No sign-in needed.

Download the Q4_K_M file (opens in a new tab)

huggingface.co/unsloth/Olmo-3.1-32B-Think-GGUF (opens in a new tab)

ollama run hf.co/unsloth/Olmo-3.1-32B-Think-GGUF:Q4_K_M

Original weights · Hugging Face

Full-precision files from Ai2, for fine-tuning, vLLM, Transformers.

huggingface.co/allenai/Olmo-3.1-32B-Think (opens in a new tab)

No sign-in needed.

Official repo · GitHub

Reference code, inference scripts and docs from Ai2. The weights themselves download from Hugging Face.

github.com/allenai/OLMo (opens in a new tab)

git clone https://github.com/allenai/OLMo

Links checked 5 days ago against Hugging Face, Ollama and GitHub.

How the S grade is scored

Every result gets up to 100 points from four parts. The points decide the letter grade.

Model qualityEditorial quality tier of the model, scaled by the quantization or precision used.32.5 / 45
Memory fitFull marks when everything sits in GPU or unified memory; fewer when part spills to system RAM.24.0 / 25
SpeedGeneration speed against a 40 tokens/s target for LLMs, or render time against a target for image and video.14.2 / 15
Memory useRewards using the memory you have, so a bigger, better variant beats a tiny one that leaves most memory idle.15.0 / 15
Total85.7 / 10085 ptsExcellent (S) · 85 of 100 points: Model quality 32.5/45, Memory fit 24.0/25, Speed 14.2/15, Memory use 15.0/15

Grade scale: S 85+ · A 72+ · B 58+ · C 42+ · D 25+ · F under 25 points.

Every quantization on the NVIDIA GeForce RTX 3090

Quant Memory needed Available Speed Fit Grade · points Download
Q2_K 15.4 GB 24.0 GB 58.5 tok/s Runs well 78 ptsGreat (A) · 78 of 100 points: Model quality 23.9/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 14.8/15 GGUF (opens in a new tab)
Q3_K_M 18.3 GB 24.0 GB 46.1 tok/s Runs well 83 ptsGreat (A) · 83 of 100 points: Model quality 28.7/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 15.0/15 GGUF (opens in a new tab)
Q4_K_M 22.2 GB 24.0 GB 35.6 tok/s Tight fit 85 ptsExcellent (S) · 85 of 100 points: Model quality 32.5/45, Memory fit 24.0/25, Speed 14.2/15, Memory use 15.0/15 GGUF (opens in a new tab)
Q5_K_M 25.4 GB 24.0 GB 16.3 tok/s Runs with offload 67 ptsGood (B) · 67 of 100 points: Model quality 33.2/45, Memory fit 10.0/25, Speed 9.6/15, Memory use 15.0/15 GGUF (opens in a new tab)
Q6_K 29.0 GB 24.0 GB 7.2 tok/s Runs with offload 65 ptsGood (B) · 65 of 100 points: Model quality 33.9/45, Memory fit 10.0/25, Speed 6.3/15, Memory use 15.0/15 GGUF (opens in a new tab)
Q8_0 36.7 GB 24.0 GB 3.1 tok/s Heavy offload 57 ptsUsable (C) · 57 of 100 points: Model quality 34.2/45, Memory fit 4.0/25, Speed 4.2/15, Memory use 15.0/15 GGUF (opens in a new tab)
F16 67.0 GB 24.0 GB — Won't run 10 ptsWon't run (F) · 10 of 100 points: Model quality 10.3/45, Memory fit 0.0/25, Speed 0.0/15, Memory use 0.0/15 GGUF · folder (opens in a new tab) Original weights (opens in a new tab)

Memory figures assume 8K of context. Longer conversations or documents grow the KV cache.

Other hardware that runs it

Estimates come from published hardware specs and our open formulas, not from running the model on your machine. Real speed varies with runtime, drivers and settings. Numbers marked “Measured” come from a linked benchmark.