Qwen3 Next 80B A3B Thinking NVFP4 — Hardware Requirements & GPU Compatibility
Chat
Qwen3 Next 80B A3B Thinking NVFP4 is NVIDIA's quantized version of Alibaba's Qwen3 Next 80B, a mixture-of-experts model with thinking capabilities and only 3 billion active parameters per token. The NVFP4 format significantly reduces memory requirements, bringing this 80B model within reach of high-end consumer hardware. The thinking mode enables explicit chain-of-thought reasoning, where the model works through problems step by step before delivering its answer. Combined with the MoE efficiency of activating just 3B parameters at a time, this model offers an unusual combination of deep reasoning and fast inference.
128.7K downloads
52 likesSpecifications
- Publisher
- NVIDIA
- Family
- Qwen
- Parameters
- 80B
- Architecture
- Qwen3NextForCausalLM
- Context Length
- 262,144 tokens
- Vocabulary Size
- 151,936
- Release Date
- 2026-02-09
- License
- Apache 2.0
Get Started
HuggingFace
How Much VRAM Does Qwen3 Next 80B A3B Thinking NVFP4 Need?
Select a quantization to see compatible GPUs below.
| Quantization | Bits | VRAM | + Context | File Size | Quality |
|---|---|---|---|---|---|
| BF16 | 16.00 | 160.4 GB | 173.2 GB | 160.00 GB | Brain floating point 16 — preferred for training |
Which GPUs Can Run Qwen3 Next 80B A3B Thinking NVFP4?
BF16 · 160.4 GBQwen3 Next 80B A3B Thinking NVFP4 (BF16) requires 160.4 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 209+ GB is recommended. Using the full 262K context window can add up to 12.8 GB, bringing total usage to 173.2 GB. No single GPU has enough memory — multi-GPU or cluster setups are needed.
Which Devices Can Run Qwen3 Next 80B A3B Thinking NVFP4?
BF16 · 160.4 GB4 devices with unified memory can run Qwen3 Next 80B A3B Thinking NVFP4, including NVIDIA DGX H100, NVIDIA DGX A100 640GB, Mac Pro M2 Ultra (192 GB).
Runs great
— Plenty of headroom
Decent
— Enough memory, may be tightRelated Models
Frequently Asked Questions
- How much VRAM does Qwen3 Next 80B A3B Thinking NVFP4 need?
Qwen3 Next 80B A3B Thinking NVFP4 requires 160.4 GB of VRAM at BF16. Full 262K context adds up to 12.8 GB (173.2 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 80B × 16 bits ÷ 8 = 160 GB
KV Cache + Overhead ≈ 0.4 GB (at 2K context + ~0.3 GB framework)
KV Cache + Overhead ≈ 13.2 GB (at full 262K context)
VRAM usage by quantization
BF16160.4 GBBF16 + full context173.2 GB- Can NVIDIA GeForce RTX 5090 run Qwen3 Next 80B A3B Thinking NVFP4?
No — Qwen3 Next 80B A3B Thinking NVFP4 requires at least 160.4 GB at BF16, which exceeds the NVIDIA GeForce RTX 5090's 32 GB of VRAM.
- Can I run Qwen3 Next 80B A3B Thinking NVFP4 on a Mac?
Qwen3 Next 80B A3B Thinking NVFP4 requires at least 160.4 GB at BF16, 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 Next 80B A3B Thinking NVFP4 locally?
Yes — Qwen3 Next 80B A3B Thinking NVFP4 can run locally on consumer hardware. At BF16 quantization it needs 160.4 GB of VRAM. Popular tools include Ollama, LM Studio, and llama.cpp.
- How fast is Qwen3 Next 80B A3B Thinking NVFP4?
At BF16, Qwen3 Next 80B A3B Thinking NVFP4 can reach ~18 tok/s on AMD Instinct MI300X. 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 ÷ 160.4 × 0.55 = ~18 tok/s
Estimated speed at BF16 (160.4 GB)
AMD Instinct MI300X~18 tok/sReal-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
- What's the download size of Qwen3 Next 80B A3B Thinking NVFP4?
At BF16, the download is about 160.00 GB.