Qwen3 32B NVFP4 — Hardware Requirements & GPU Compatibility
Chat
Qwen3 32B NVFP4 is NVIDIA's NVFP4-quantized version of Alibaba's dense Qwen3 32B model, reduced to approximately 17.2 billion parameters of effective memory usage. Unlike the MoE variants, this is a traditional dense model where all parameters contribute to every token, often yielding more consistent output quality. Qwen3 32B has earned a strong reputation as one of the best models in its size class, and NVIDIA's NVFP4 quantization makes it accessible on a broader range of GPUs. If you prefer the predictability of a dense architecture over MoE's efficiency trade-offs, this is the variant to choose.
76.7K downloads
13 likesSpecifications
- Publisher
- NVIDIA
- Family
- Qwen
- Parameters
- 17.2B
- Architecture
- Qwen3ForCausalLM
- Context Length
- 40,960 tokens
- Vocabulary Size
- 151,936
- Release Date
- 2025-09-09
- License
- Apache 2.0
Get Started
HuggingFace
How Much VRAM Does Qwen3 32B NVFP4 Need?
Select a quantization to see compatible GPUs below.
| Quantization | Bits | VRAM | + Context | File Size | Quality |
|---|---|---|---|---|---|
| Q4_K_M | 4.80 | 10.9 GB | 17.3 GB | 10.30 GB | 4-bit medium quantization — most popular sweet spot |
| Q5_0 | 5.00 | 11.4 GB | 17.7 GB | 10.72 GB | 5-bit legacy quantization |
| Q5_K_M | 5.70 | 12.9 GB | 19.2 GB | 12.23 GB | 5-bit medium quantization — good quality/size tradeoff |
| Q6_K | 6.60 | 14.8 GB | 21.2 GB | 14.16 GB | 6-bit quantization, very good quality |
| Q8_0 | 8.00 | 17.8 GB | 24.2 GB | 17.16 GB | 8-bit quantization, near-lossless |
Which GPUs Can Run Qwen3 32B NVFP4?
Q4_K_M · 10.9 GBQwen3 32B NVFP4 (Q4_K_M) requires 10.9 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 15+ GB is recommended. Using the full 41K context window can add up to 6.4 GB, bringing total usage to 17.3 GB. 27 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti, NVIDIA GeForce RTX 5080.
Runs great
— Plenty of headroom
Decent
— Enough VRAM, may be tightWhich Devices Can Run Qwen3 32B NVFP4?
Q4_K_M · 10.9 GB27 devices with unified memory can run Qwen3 32B NVFP4, including NVIDIA DGX H100, NVIDIA DGX A100 640GB, Mac Mini M4 (16 GB).
Runs great
— Plenty of headroomRelated Models
Frequently Asked Questions
- How much VRAM does Qwen3 32B NVFP4 need?
Qwen3 32B NVFP4 requires 10.9 GB of VRAM at Q4_K_M, or 17.8 GB at Q8_0. Full 41K context adds up to 6.4 GB (17.3 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 17.2B × 4.8 bits ÷ 8 = 10.3 GB
KV Cache + Overhead ≈ 0.6 GB (at 2K context + ~0.3 GB framework)
KV Cache + Overhead ≈ 7 GB (at full 41K context)
VRAM usage by quantization
Q4_K_M10.9 GBQ4_K_M + full context17.3 GB- What's the best quantization for Qwen3 32B NVFP4?
For Qwen3 32B NVFP4, Q4_K_M (10.9 GB) offers the best balance of quality and VRAM usage. Q5_0 (11.4 GB) provides better quality if you have the VRAM.
VRAM requirement by quantization
Q4_K_M ★10.9 GB~89%Q5_011.4 GB~90%Q5_K_M12.9 GB~92%Q6_K14.8 GB~95%Q8_017.8 GB~99%★ Recommended — best balance of quality and VRAM usage.
- Can I run Qwen3 32B NVFP4 on a Mac?
Qwen3 32B NVFP4 requires at least 10.9 GB at Q4_K_M, 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 Qwen3 32B NVFP4 locally?
Yes — Qwen3 32B NVFP4 can run locally on consumer hardware. At Q4_K_M quantization it needs 10.9 GB of VRAM. Popular tools include Ollama, LM Studio, and llama.cpp.
- How fast is Qwen3 32B NVFP4?
At Q4_K_M, Qwen3 32B NVFP4 can reach ~267 tok/s on AMD Instinct MI300X. On NVIDIA GeForce RTX 4090: ~60 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 ÷ 10.9 × 0.55 = ~267 tok/s
Estimated speed at Q4_K_M (10.9 GB)
AMD Instinct MI300X~267 tok/sNVIDIA GeForce RTX 4090~60 tok/sNVIDIA H100 SXM~199 tok/sAMD Instinct MI250X~165 tok/sReal-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
- What's the download size of Qwen3 32B NVFP4?
At Q4_K_M, the download is about 10.30 GB. The full-precision Q8_0 version is 17.16 GB.