QW

Qwen.Qwen3.5 397B A17B

QwenGGUF by DevQuasarUpdated 2026-02-17
Download on Hugging Face
Parameters396B
Context256K
Layers60
Architectureqwen35moe
Smallest quant115.8 GB
Quick answer

Qwen.Qwen3.5 397B A17B needs about 242.1 GB of memory at Q4_K_M (a 240.6 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 8 quants, from 115.8 GB (IQ2_XS) to 421.5 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 4.0 GB.

Qwen.Qwen3.5 397B A17B – a large multimodal LLM with 396B parameters and 17B active experts

Qwen.Qwen3.5 397B A17B is a 396B‑parameter model with 17B active mixture‑of‑expert parameters. It supports reasoning, tools, and vision tasks. The model is available in multiple GGUF quantizations ranging from IQ2_XS to Q8_0.

Good at

  • Reasoning and complex problem solving
  • Tool use and API integration
  • Vision‑language tasks

Limits

  • Requires large VRAM for full precision
  • No native vision support in some runtimes
  • License is apache‑2.0

Best for

  • Enterprise inference on high‑end GPUs
  • Research on multimodal capabilities
  • Tool‑augmented applications

Who it’s for. Developers and researchers needing a powerful multimodal LLM for advanced reasoning, tool integration, and vision tasks. The model is suitable for environments with sufficient GPU memory and compliance with the apache‑2.0 license.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
IQ2_XS115.8 GB (8 parts) 117.4 GB
Too big Get file →
Q2_K144.2 GB (10 parts) 145.7 GB
Too big Get file →
Q3_K_M189.5 GB (13 parts) 191.0 GB
Too big Get file →
MXFP4_MOE216.2 GB (14 parts) 217.7 GB
Too big Get file →
Q4_K_M240.6 GB (16 parts) 242.1 GB
Too big Get file →
Q5_K_M281.9 GB (19 parts) 283.4 GB
Too big Get file →
Q6_K325.5 GB (23 parts) 327.1 GB
Too big Get file →
Q8_0421.5 GB (31 parts) 423.0 GB
Too big Get file →

For 8 GB VRAM, the IQ2_XS quant (115.85 GB file, 117.4 GB memory at 8k context) is the smallest option. For 16 GB VRAM, Q2_K (144.21 GB file, 145.7 GB memory) or Q3_K_M (189.5 GB file, 191 GB memory) are viable. Higher quants such as Q4_K_M or Q5_K_M provide better quality at the cost of more memory.

What GPU do you need for Qwen.Qwen3.5 397B A17B?

These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.

  • 4 GB VRAM: no quant fits at an 8K context.
  • 6 GB VRAM: no quant fits at an 8K context.
  • 8 GB VRAM: no quant fits at an 8K context.
  • 12 GB VRAM: no quant fits at an 8K context.
  • 16 GB VRAM: no quant fits at an 8K context.
  • 24 GB VRAM: no quant fits at an 8K context.
  • 32 GB VRAM: no quant fits at an 8K context.
  • 48 GB VRAM: no quant fits at an 8K context.

Which Qwen.Qwen3.5 397B A17B quant should you choose?

Q4_K_M is the usual starting point for Qwen.Qwen3.5 397B A17B: it balances file size and quality well. Here it is a 240.6 GB file.

The Q8_0 file is 421.5 GB, about 1.8 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.

If memory is tight, the smallest file is IQ2_XS at 115.8 GB, with the quality trade-offs that come with fewer bits. Use it only when memory is very tight. Expect a clear drop in quality, which is usually worse on small models and easier to live with on large ones.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architectureqwen35moe
Parameters396B (approximate, mixture of experts)
Active parameters17B
Layers60
Trained context262,144 tokens
KV cache (fp16)120 KB per token

Source

Quantized byDevQuasar
FamilyQwen
Licenseapache-2.0
Downloads (HF)568
Last updated2026-02-17

Frequently asked questions about Qwen.Qwen3.5 397B A17B

How much VRAM does Qwen.Qwen3.5 397B A17B need?

About 242.1 GB at Q4_K_M with an 8K context window. That is the 240.6 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.

Can I run Qwen.Qwen3.5 397B A17B on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 117.4 GB, so you would need a larger GPU or CPU offloading.

What is the file size of Qwen.Qwen3.5 397B A17B GGUF?

It depends on the quant: IQ2_XS is 115.8 GB, Q2_K is 144.2 GB, Q3_K_M is 189.5 GB, MXFP4_MOE is 216.2 GB, Q4_K_M is 240.6 GB, Q5_K_M is 281.9 GB and 2 more.

What context length does Qwen.Qwen3.5 397B A17B support?

The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 1.0 GB at 8K and 4.0 GB at 32K.

Where can I download Qwen.Qwen3.5 397B A17B GGUF?

The files are on Hugging Face in the DevQuasar/Qwen.Qwen3.5-397B-A17B-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

What license does Qwen.Qwen3.5 397B A17B use?

Hugging Face lists the license as apache-2.0. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.

Can I run Qwen.Qwen3.5 397B A17B on 8 GB VRAM?

Only the smallest quant, IQ2_XS, may fit, but memory usage is still above 8 GB. A larger GPU is recommended.

Does the model support vision tasks?

Yes, the model has detected vision ability, but some runtimes may require additional configuration.

What license applies to this model?

The model is released under the apache‑2.0 license.

Other GGUFs of Qwen3.5-397B-A17B

Similar models

Browse all →

More from DevQuasar

All by DevQuasar →

More Qwen models

All Qwen →

Other large models

All large models →