Qwen3 32B (Q5_K_M) — 23.4 GBon NVIDIA RTX 3090

Q5_K_M NVIDIA RTX 3090

Overview

Qwen3 32B is a 32B parameter dense language model by Qwen, with code, multilingual, thinking, tool-calls capabilities. It supports a context window of up to 40,960 tokens.

Qwen3 32B is a 32-billion-parameter dense transformer from Alibaba's Qwen team, combining thinking capabilities with strong code generation, tool calling, and multilingual support. It occupies a mid-range parameter class that balances reasoning depth with practical deployment requirements, outperforming many larger models on math and logic benchmarks. The model supports 14 languages including English, Chinese, and Arabic. With a 40K context window and flash attention, it fits on a single high-end GPU at Q4 quantization for self-hosted inference.

At Q5_K_M quantization (medium quality tier), the model weighs 21.62 GB. This fits within the 24 GB of VRAM on NVIDIA RTX 3090, enabling full GPU inference.

The NVIDIA GeForce RTX 3090 is a consumer GPU with 24 GB of GDDR6X VRAM and 936 GB/s memory bandwidth. It provides 35.6 FP16 TFLOPS, offering solid performance for local LLM inference at a lower cost than newer cards. It runs quantized models up to 20B parameters well. A practical choice for budget-minded developers and home lab enthusiasts.

Hardware Requirements

Model size	21.62 GB
VRAM available	24 GB
VRAM used	23.4 GB
GPU layers	64 / 64
Context size	3,507
Backend	cuda13
Flash attention	Yes

Performance Notes

Shorter conversations

Context size has been reduced due to memory constraints. This may affect the model's ability to handle long conversations.

Deploy

Server Local

Prerequisites

Ensure your GPU nodes are prepared with the NVIDIA container toolkit:

ansible-playbook prositronic.infra.nvidia_container_toolkit

Command

helmfile --state-values-file <(curl -s https://www.prositronic.eu/values/qwen3-32b/q5_k_m/nvidia-rtx3090.yaml) apply

Generated values.yaml

/values/qwen3-32b/q5_k_m/nvidia-rtx3090.yaml

Loading values…

Install llama.cpp

Install llama.cpp from the official build instructions for your platform and backend.

Download Model

curl -L -o qwen3-32b.gguf "https://huggingface.co/unsloth/Qwen3-32B-GGUF/resolve/main/Qwen3-32B-Q5_K_M.gguf"

Start Server

llama-server \
  -m qwen3-32b.gguf \
  --n-gpu-layers 64 \
  --ctx-size 3507 \
  --flash-attn

Verify

curl http://localhost:8080/health

Explore

• Configure a deployment
• Browse models
• Browse hardware

Other quantizations

• Qwen3 32B (Q8_0)
• Qwen3 32B (Q4_K_M)

Other hardware

• Qwen3 32B on CPU Only
• Qwen3 32B on NVIDIA A100 80GB
• Qwen3 32B on NVIDIA H100

Frequently Asked Questions

How much VRAM does Qwen3 32B (Q5_K_M) need?

The Q5_K_M quantization of Qwen3 32B requires 21.62 GB. All 64 layers fit in the 24 GB of VRAM available on NVIDIA RTX 3090, enabling full GPU acceleration.

Can I run Qwen3 32B on NVIDIA RTX 3090?

Yes. NVIDIA RTX 3090 provides 24 GB of VRAM, which is enough to run Qwen3 32B (Q5_K_M) with all layers on the GPU for optimal performance.

What is quantization?

Quantization reduces a model's numerical precision from its original floating-point format to a more compact representation. This shrinks the file size and VRAM footprint, making it possible to run large models on consumer hardware. The trade-off is a small reduction in output quality. Q5_K_M compresses Qwen3 32B from its original size down to 21.62 GB.

What quantization should I choose for Qwen3 32B?

Q5_K_M is a medium-quality quantization. Higher-quality quants (Q8, Q6) preserve more model accuracy but need more VRAM. Lower quants (Q4, Q3, Q2) reduce VRAM usage at the cost of some quality. Choose based on your available hardware and quality requirements.

What is flash attention and why is it enabled?

Flash attention is a memory-efficient algorithm that speeds up the attention mechanism in transformer models. It reduces VRAM usage during inference by avoiding the materialisation of the full attention matrix. For Qwen3 32B on NVIDIA RTX 3090, flash attention is enabled to maximise context length and throughput within the available 24 GB of VRAM.

How do I run Qwen3 32B (Q5_K_M) with Ollama?

Run ollama run qwen3:32b-q5_k_m to start Qwen3 32B (Q5_K_M). Ollama handles downloading the model weights automatically on first run.

Last updated: March 5, 2026