Can NVIDIA GeForce RTX 4090 run QwQ 32B?

Yes, at Q5_K_M (23.6 GB) or lower. Higher quantizations like Q6_K (27.2 GB) exceed the NVIDIA GeForce RTX 4090's 24 GB.

What's the best quantization for QwQ 32B?

For QwQ 32B, Q4_K_M (20.0 GB) offers the best balance of quality and VRAM usage. Q5_K_M (23.6 GB) provides better quality if you have the VRAM. The smallest option is Q2_K at 14.4 GB.

Can I run QwQ 32B on a Mac?

QwQ 32B requires at least 14.4 GB at Q2_K, which exceeds the unified memory of most consumer Macs. You would need a Mac Studio or Mac Pro with a high-memory configuration.

Can I run QwQ 32B locally?

Yes — QwQ 32B can run locally on consumer hardware. At Q4_K_M quantization it needs 20.0 GB of VRAM. Popular tools include Ollama, LM Studio, and llama.cpp.

At Q4_K_M, QwQ 32B can reach ~146 tok/s on AMD Instinct MI300X. On NVIDIA GeForce RTX 4090: ~33 tok/s. Speed depends mainly on GPU memory bandwidth. Real-world results typically within ±20%.

What's the download size of QwQ 32B?

At Q4_K_M, the download is about 19.20 GB. The full-precision Q8_0 version is 32.00 GB. The smallest option (Q2_K) is 13.60 GB.

Alibaba·QwQ·Qwen2ForCausalLM

QwQ 32B — Hardware Requirements & GPU Compatibility

ChatReasoning

QwQ 32B is a 32-billion parameter reasoning-focused model from Alibaba Cloud's Qwen family. Unlike standard chat models, QwQ is specifically optimized for step-by-step logical reasoning, complex problem solving, and mathematical tasks. It employs extended chain-of-thought processing, generating detailed internal reasoning before producing final answers, which significantly improves accuracy on challenging analytical problems. The model requires a GPU with at least 24GB of VRAM for quantized inference and delivers reasoning performance competitive with much larger models. It is particularly well suited for users who need strong analytical capabilities for math, science, coding logic, and multi-step problem solving. Released under the Apache 2.0 license.

52.6K downloads 2.9K likesMar 202541K context

Specifications

Publisher: Alibaba
Family: QwQ
Parameters: 32B
Architecture: Qwen2ForCausalLM
Context Length: 40,960 tokens
Vocabulary Size: 152,064
Release Date: 2025-03-11
License: Apache 2.0

Get Started

HuggingFace

Qwen/QwQ-32B

How Much VRAM Does QwQ 32B Need?

Select a quantization to see compatible GPUs below.

Quantization	Bits	VRAM	+ Context	File Size	Quality
Q2_K	3.40	14.4 GB	24.6 GB	13.60 GB	2-bit quantization with K-quant improvements
Q3_K_M	3.90	16.4 GB	26.6 GB	15.60 GB	3-bit medium quantization
Q3_K_L	4.10	17.2 GB	27.4 GB	16.40 GB	3-bit large quantization
Q4_K_M	4.80	20.0 GB	30.2 GB	19.20 GB	4-bit medium quantization — most popular sweet spot
Q5_K_M	5.70	23.6 GB	33.8 GB	22.80 GB	5-bit medium quantization — good quality/size tradeoff
Q6_K	6.60	27.2 GB	37.4 GB	26.40 GB	6-bit quantization, very good quality
Q8_0	8.00	32.8 GB	43.0 GB	32.00 GB	8-bit quantization, near-lossless

Which GPUs Can Run QwQ 32B?

Q4_K_M · 20.0 GB

Show professional

QwQ 32B (Q4_K_M) requires 20.0 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 41K context window can add up to 10.2 GB, bringing total usage to 30.2 GB. 5 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti.

Runs great

— Plenty of headroom

NVIDIA GeForce RTX 5090~58 tok/s

Decent

— Enough VRAM, may be tight

NVIDIA GeForce RTX 3090 Ti~33 tok/s NVIDIA GeForce RTX 4090~33 tok/s NVIDIA GeForce RTX 3090~30 tok/s AMD Radeon RX 7900 XTX~26 tok/s

Which Devices Can Run QwQ 32B?

Q4_K_M · 20.0 GB

21 devices with unified memory can run QwQ 32B, including NVIDIA DGX H100, NVIDIA DGX A100 640GB, Mac Mini M4 Pro (24 GB).

Related Models

Derivatives (5)

AStar Thought QwQ 32B Openbuddy Thinker 32B V26 Preview QwQ 32B Abliterated QwQ 32B FP8 Dynamic Seneca Cybersecurity LLM X QwQ 32B Q4 Medium Version

Frequently Asked Questions

How much VRAM does QwQ 32B need?: QwQ 32B requires 20.0 GB of VRAM at Q4_K_M, or 32.8 GB at Q8_0. Full 41K context adds up to 10.2 GB (30.2 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 32B × 4.8 bits ÷ 8 = 19.2 GB
KV Cache + Overhead ≈ 0.8 GB (at 2K context + ~0.3 GB framework)
KV Cache + Overhead ≈ 11 GB (at full 41K context)
VRAM usage by quantization
Q4_K_M
20.0 GB
Q4_K_M + full context
30.2 GB
Learn more about VRAM estimation →
Can NVIDIA GeForce RTX 4090 run QwQ 32B?: Yes, at Q5_K_M (23.6 GB) or lower. Higher quantizations like Q6_K (27.2 GB) exceed the NVIDIA GeForce RTX 4090's 24 GB.
What's the best quantization for QwQ 32B?: For QwQ 32B, Q4_K_M (20.0 GB) offers the best balance of quality and VRAM usage. Q5_K_M (23.6 GB) provides better quality if you have the VRAM. The smallest option is Q2_K at 14.4 GB.
VRAM requirement by quantization
Q2_K
14.4 GB~75%
Q3_K_L
17.2 GB~86%
Q4_K_M ★
20.0 GB~89%
Q5_K_M
23.6 GB~92%
Q6_K
27.2 GB~95%
Q8_0
32.8 GB~99%
★ Recommended — best balance of quality and VRAM usage.
Learn more about quantization →
Can I run QwQ 32B on a Mac?: QwQ 32B requires at least 14.4 GB at Q2_K, which exceeds the unified memory of most consumer Macs. You would need a Mac Studio or Mac Pro with a high-memory configuration.
Can I run QwQ 32B locally?: Yes — QwQ 32B can run locally on consumer hardware. At Q4_K_M quantization it needs 20.0 GB of VRAM. Popular tools include Ollama, LM Studio, and llama.cpp.
How fast is QwQ 32B?: At Q4_K_M, QwQ 32B can reach ~146 tok/s on AMD Instinct MI300X. 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: AMD Instinct MI300X → 5300 ÷ 20.0 × 0.55 = ~146 tok/s
Estimated speed at Q4_K_M (20.0 GB)
AMD Instinct MI300X
~146 tok/s
NVIDIA GeForce RTX 4090
~33 tok/s
NVIDIA H100 SXM
~109 tok/s
AMD Instinct MI250X
~90 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 QwQ 32B?: At Q4_K_M, the download is about 19.20 GB. The full-precision Q8_0 version is 32.00 GB. The smallest option (Q2_K) is 13.60 GB.