QW

GRAFT 50B A3B

QwenGGUF by SmallAICreatorUpdated 2026-10-21
Download on Hugging Face
Parameters50.7B
Context256K
Layers48
Architectureqwen3next
Smallest quant30.9 GB
Quick answer

GRAFT 50B A3B needs about 32.2 GB of memory at Q4_K_M (a 30.9 GB file) with an 8K context window, so it fits on a 48 GB GPU. Longer context needs more: at 32K tokens the KV cache alone adds about 3.2 GB.

GRAFT 50B A3B – a 50.7B parameter MoE chat model with 3B active per token

GRAFT 50B A3B is a 50.7B parameter mixture‑of‑experts model based on Qwen3‑Next‑80B‑A3B‑Instruct. Only 3.2B parameters are active per token, enabling inference speed similar to a 3B model while retaining large‑scale knowledge. The model supports tool calling and uses the ChatML template.

Good at

  • Tool‑aware chat
  • Ultra‑long context up to 262144 tokens
  • Efficient inference with 3B active parameters

Limits

  • Requires large VRAM – ~32.2 GB at 8 k context
  • No vision or audio modalities
  • Apache 2.0 license restricts commercial redistribution

Best for

  • Building lightweight agents that need large context
  • Research on MoE efficiency
  • Deploying chatbots with tool integration

Who it’s for. Developers and researchers who need a high‑capacity chat model that runs at the speed of a 3B model. The model is suitable for applications requiring extensive context handling and tool usage.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q4_K_M30.9 GB 32.2 GB
Too big Get file →

The only available quant is Q4_K_M, 30.86 GB in size. It requires approximately 32.2 GB of VRAM at 8 k context. For systems with less VRAM, consider using a higher‑quantization model if available, but no other quants are listed.

What GPU do you need for GRAFT 50B A3B?

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: the best quant that fits is Q4_K_M (30.9 GB file, about 32.2 GB in use).

Which GRAFT 50B A3B quant should you choose?

Q4_K_M is the usual starting point for GRAFT 50B A3B: it balances file size and quality well. Here it is a 30.9 GB file.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architectureqwen3next
Parameters50.7B (approximate, mixture of experts)
Active parameters3B
Layers48
Trained context262,144 tokens
KV cache (fp16)96 KB per token

Source

Quantized bySmallAICreator
FamilyQwen
Licenseapache-2.0
Downloads (HF)1,370
Last updated2026-10-21

Frequently asked questions about GRAFT 50B A3B

How much VRAM does GRAFT 50B A3B need?

About 32.2 GB at Q4_K_M with an 8K context window. That is the 30.9 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 GRAFT 50B A3B on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 32.2 GB, so you would need a larger GPU or CPU offloading.

What is the file size of GRAFT 50B A3B GGUF?

It depends on the quant: Q4_K_M is 30.9 GB.

What context length does GRAFT 50B A3B support?

The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 0.8 GB at 8K and 3.2 GB at 32K.

Where can I download GRAFT 50B A3B GGUF?

The files are on Hugging Face in the SmallAICreator/GRAFT-50B-A3B-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 GRAFT 50B A3B 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 GRAFT 50B A3B on 8 GB VRAM?

No. The Q4_K_M quant requires about 32.2 GB of VRAM at 8 k context, which exceeds 8 GB.

Does GRAFT 50B A3B support vision or audio?

No. The model does not include vision or audio modalities.

What license applies to GRAFT 50B A3B?

The model is licensed under Apache 2.0, allowing modification and redistribution under its terms.

Other GGUFs of Qwen3-Next-80B-A3B-Instruct

Similar models

Browse all →

More from SmallAICreator

All by SmallAICreator →

More Qwen models

All Qwen →

Other large models

All large models →