Qwen3.6 28B — Hardware Requirements & GPU Compatibility
Chat
Qwen3.6 28B is a 28.2B-parameter open language model from 0xSero in the Qwen 3.6 family. It supports a context window of up to 262,144 tokens. At Q4_K_M it needs about 17.33 GB of VRAM — see which GPUs and Macs can run it below.
Runs great on a 24 GB GPU
Runs from 16 GB (Q3_K_M · 128K)Best on 24 GB (Q4_K_M · ~184 tok/s · 256K)Macs from 16 GB (Q2_K · ~53 tok/s)
Apache 2.0
Reference GPU · RTX 4090 24 GB
Estimates, ±20% · how we estimate
Run Qwen3.6 28B
One recommended way. Every other method, OS and quant is in the step-by-step guide.
Recommended
vllm serve 0xSero/Qwen3.6-28B --max-model-len 16384vLLM repo with its own quantisation
Download: Weights on Hugging Face
Other ways
each opens step-by-stepRun in cloud
Fits on RTX 3090 (24 GB) (6 GB headroom) · Q4_K_M
- Generation speed
- ~176 tok/s
- generation speed
- Cost per 1M output tokens
- $0.32
- per 1M output tokens
How Much VRAM Does Qwen3.6 28B Need?
Select a quantization level; the listed hardware adheres to it.
| Quantization | VRAM | +32K | On RTX 4090 (reference) | |
|---|---|---|---|---|
| 12.4 GB | 12.6 GB | 2-bit · K-quant — 2-bit quantization with K-quant improvements | Fits: 256K context · ~222 tok/s | |
| 14.2 GB | 14.4 GB | 3-bit medium — 3-bit medium quantization | Fits: 256K context · ~207 tok/s | |
| 17.3 GB | 17.6 GB | 4-bit medium — 4-bit medium quantization — most popular sweet spot | Fits: 256K context · ~184 tok/s | |
| 20.5 GB | 20.8 GB | 5-bit medium — 5-bit medium quantization — good quality/size tradeoff | Fits: 256K context · ~166 tok/s | |
| 23.7 GB | 23.9 GB | 6-bit, very good — 6-bit quantization, very good quality | Fits: 32K context · ~151 tok/s | |
| 28.6 GB | 28.9 GB | 8-bit, near-lossless — 8-bit quantization, near-lossless | Does not fit: needs 28.7 GB | |
| 56.9 GB | 57.1 GB | 16-bit · bfloat — Brain floating point 16 — preferred for training | Does not fit: needs 56.9 GB |
Selected Q4_K_M
4-bit medium quantization17.4GBat 16K context
- Weights
- 17.0 GB
- KV cache
- 0.2 GB
- Overhead
- 0.3 GB
Fits RTX 4090 (reference) · 6.6 GB free · ~184 tok/s
≥50 tok/s needs ≥183 GB/s · RTX 5090 ~249, RTX 3090 Ti ~184
Which Hardware Can Run Qwen3.6 28B?
Q4_K_M · 17.3 GBQwen3.6 28B (Q4_K_M) requires 17.3 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 23+ GB is recommended. Using the full 262K context window can add up to 2.6 GB, bringing total usage to 19.9 GB. 8 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti. 122 devices with unified memory can run Qwen3.6 28B, including Mac Studio (M3 Ultra, 256GB), Mac Studio (M3 Ultra, 512GB), Mac Mini M5 Pro (24 GB).
20 GB
fits · 2.6 GB free · up to 256K ctx
1 GPU
RX 7900 XT~81Simulate output on RX 7900 XT32–40 GB
fits · 14.6 GB free · up to 256K ctx
1 GPU · 3 Mac classes · 4 device families
RTX 5090~249Simulate output on RTX 5090M5 Max36 GB~69Simulate output on Mac Studio M5 Max (36 GB)M532 GB~36Simulate output on MacBook Air 13" M5 (32 GB)Orin~55Simulate output on NVIDIA Jetson AGX Orin 32GBSnapdragon X Elite (X1E-80-100)~36Simulate output on Snapdragon X Elite Copilot+ PC
By VRAM tier · typical GPUs and Macs at Q4_K_M
Alternatives to Qwen3.6 28B
Qwen 3.6 family · 15 models →Similar models
6| Model | Score | VRAM | tok/s | Why it is similar |
|---|---|---|---|---|
| Qwen3.8 27BAlibaba | 122.1 | 17.9 GB | ~38 | same size |
| Qwen3.5 27BAlibaba | 103.8 | 17.9 GB | ~38 | same size |
| Hy Memory 30B A3BTencent | — | 19.1 GB | ~184 | same size |
| Muse Glimmer 30BMeta | — | 19.6 GB | ~36 | same size |
| Hy MT2 30B A3BTencent | — | 19.1 GB | ~170 | same size |
| Gemma 4 26B A4BGoogle | — | 16.9 GB | ~153 | same size |
Specifications
- Publisher
- 0xSero
- Family
- Qwen 3.6
- Parameters
- 28.2B
- Active parameters
- 3.5B per token
- Architecture
- Qwen3_5MoeForCausalLM
- Context Length
- 262,144 tokens
- Vocabulary Size
- 248,320
- Release Date
- 2026-04-18
- License
- Apache 2.0
Frequently Asked Questions
How much VRAM does Qwen3.6 28B need?
Qwen3.6 28B requires 17.3 GB of VRAM at Q4_K_M, or 56.9 GB at BF16. Full 262K context adds up to 2.6 GB (19.9 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 28.2B × 4.8 bits ÷ 8 = 16.9 GB
KV Cache + Overhead ≈ 0.4 GB (at 2K context + ~0.3 GB framework)
Fit ratings and hardware model lists check this model with room for a 16K-token context, which needs a little more memory.
KV Cache + Overhead ≈ 3 GB (at full 262K context)
VRAM usage by quantization
Can NVIDIA GeForce RTX 4090 run Qwen3.6 28B?
Yes, at Q6_K (23.7 GB) or lower. Higher quantizations like Q8_0 (28.6 GB) exceed the NVIDIA GeForce RTX 4090's 24 GB.
What's the best quantization for Qwen3.6 28B?
For Qwen3.6 28B, Q4_K_M (17.3 GB) offers the best balance of quality and VRAM usage. Q5_K_M (20.5 GB) provides better quality if you have the VRAM. The smallest option is Q2_K at 12.4 GB.
VRAM requirement by quantization
★ Recommended — best balance of quality and VRAM usage.
Can I run Qwen3.6 28B on a Mac?
Yes — Mac Mini M2 Pro (16 GB) and 133 other Macs can run Qwen3.6 28B, the smallest at Q2_K. Apple Silicon uses unified memory, so the model shares RAM with the system. At Q4_K_M you need at least 17.3 GB of usable unified memory (RAM minus macOS overhead), from Mac Mini M5 Pro (24 GB).
Can I run Qwen3.6 28B locally?
Yes — Qwen3.6 28B can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.
How fast is Qwen3.6 28B?
At Q4_K_M, Qwen3.6 28B can reach ~119 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~184 tok/s. Speed depends mainly on GPU memory bandwidth. Real-world results typically within ±20%.
tok/s = 1000 ÷ (active GB ÷ (bandwidth GB/s × efficiency) × 1000 + layers × routing ms)
Mixture-of-Experts: just the active experts get read for each token, along with a fixed routing cost per layer.
Example: NVIDIA B200 → 2.1 GB active ÷ (8000 × 0.65) = 0.407 ms, plus 40 layers × 0.055 ms = 2.200 ms, so 1000 ÷ 2.607 ms = ~384 tok/s
Estimated speed at Q4_K_M (17.3 GB)
Real-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
What's the download size of Qwen3.6 28B?
At Q4_K_M, the download is about 16.94 GB. The full-precision BF16 version is 56.48 GB. The smallest option (Q2_K) is 12.00 GB.
Which GPUs can run Qwen3.6 28B?
8 consumer GPUs can run Qwen3.6 28B at Q4_K_M. Top options include NVIDIA GeForce RTX 5090, AMD Radeon RX 7900 XT, AMD Radeon RX 7900 XTX. 1 GPU has plenty of headroom for comfortable inference.
Which devices can run Qwen3.6 28B?
122 devices with unified memory can run Qwen3.6 28B at Q4_K_M (17.3 GB), including AMD Ryzen AI 9 HX 370 (Strix Point) Laptop, ASUS Ascent GX10, Beelink GTR9 Pro (Ryzen AI Max+ 395, 128 GB), Framework Desktop (Ryzen AI Max+ 395, 128 GB). Apple Silicon Macs use unified memory shared between CPU and GPU, making them well-suited for local LLM inference.