GR

Grug 3b

GrugGGUF by bartowskiUpdated 2026-07-28
Download on Hugging Face
Parameters4.17B
Context256K
Layers22
Architecturenanbeige
Smallest quant0.0 GB
Quick answer

Grug 3b needs about 3.9 GB of memory at Q4_K_M (a 2.7 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 25 quants, from 0.0 GB to 8.3 GB (BF16). Longer context needs more: at 32K tokens the KV cache alone adds about 3.0 GB.

Grug 3b – a 4.17B parameter model focused on reasoning and tool use

Grug 3b is a 4.17B parameter model built on the Nanbeige architecture. It is optimized for reasoning tasks and tool integration, with a training context of 262144 tokens. The model is released under the Apache‑2.0 license.

Good at

  • Reasoning with tool calls
  • Balanced math and factual performance
  • Short, concise answers

Limits

  • Large context requires high VRAM
  • No vision capabilities
  • Requires 8k context for memory estimates

Best for

  • Chat assistants needing reasoning
  • Tool‑enabled workflows
  • Educational math tutoring

Who it’s for. Developers and researchers who need a lightweight reasoning model that can be run on consumer hardware. The model is suitable for integration into local LLM deployments that support llama.cpp or similar runtimes.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
—0.0 GB 1.2 GB
Fits Get file →
Q2_K1.8 GB 3.1 GB
Fits Get file →
IQ2_M1.9 GB 3.1 GB
Fits Get file →
IQ3_XXS1.9 GB 3.1 GB
Fits Get file →
IQ3_XS2.0 GB 3.3 GB
Fits Get file →
Q3_K_S2.1 GB 3.3 GB
Fits Get file →
IQ3_M2.2 GB 3.4 GB
Fits Get file →
Q3_K_M2.2 GB 3.5 GB
Fits Get file →
Q3_K_L2.3 GB 3.6 GB
Fits Get file →
Q2_K_L2.3 GB 3.6 GB
Fits Get file →
IQ4_XS2.4 GB 3.7 GB
Fits Get file →
Q4_02.5 GB 3.8 GB
Fits Get file →
IQ4_NL2.5 GB 3.8 GB
Fits Get file →
Q4_K_S2.6 GB 3.8 GB
Fits Get file →
Q4_K_M2.7 GB 3.9 GB
Fits Get file →
Q4_12.7 GB 4.0 GB
Fits Get file →
Q3_K_XL2.8 GB 4.0 GB
Fits Get file →
Q5_K_S3.0 GB 4.2 GB
Fits Get file →
Q4_K_L3.1 GB 4.3 GB
Fits Get file →
Q5_K_M3.1 GB 4.3 GB
Fits Get file →
Q5_K_L3.4 GB 4.6 GB
Fits Get file →
Q6_K3.6 GB 4.8 GB
Fits Get file →
Q6_K_L3.8 GB 5.1 GB
Fits Get file →
Q8_04.4 GB 5.7 GB
Fits Get file →
BF168.3 GB 9.6 GB
Too big Get file →

For 8 GB VRAM, use Q4_0 (file size 2.54 GB, estimated 3.8 GB memory). If 12 GB VRAM is available, Q4_K_M (2.68 GB, 3.9 GB memory) offers a good balance. For 16 GB VRAM, Q5_K_L (3.39 GB, 4.6 GB memory) provides higher quality while staying within limits.

What GPU do you need for Grug 3b?

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: the best quant that fits is Q3_K_XL (2.8 GB file, about 4.0 GB in use).
  • 6 GB VRAM: the best quant that fits is Q8_0 (4.4 GB file, about 5.7 GB in use).
  • 8 GB VRAM: the best quant that fits is Q8_0 (4.4 GB file, about 5.7 GB in use).
  • 12 GB VRAM: the best quant that fits is BF16 (8.3 GB file, about 9.6 GB in use).

Which Grug 3b quant should you choose?

Q4_K_M is the usual starting point for Grug 3b: it balances file size and quality well. Here it is a 2.7 GB file.

The Q8_0 file is 4.4 GB, about 1.7 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 the first one in the table at 0.0 GB, with the quality trade-offs that come with fewer bits.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturenanbeige
Parameters4.17B
Layers22
Trained context262,144 tokens
KV cache (fp16)88 KB per token

Source

Quantized bybartowski
FamilyGrug
Licenseapache-2.0
Downloads (HF)1,594
Last updated2026-07-28

Frequently asked questions about Grug 3b

How much VRAM does Grug 3b need?

About 3.9 GB at Q4_K_M with an 8K context window. That is the 2.7 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 Grug 3b on an 8 GB GPU?

Yes, with Q8_0: the 4.4 GB file needs about 5.7 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of Grug 3b GGUF?

It depends on the quant: file is 0.0 GB, Q2_K is 1.8 GB, IQ2_M is 1.9 GB, IQ3_XXS is 1.9 GB, IQ3_XS is 2.0 GB, Q3_K_S is 2.1 GB and 19 more.

What context length does Grug 3b support?

The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 0.7 GB at 8K and 3.0 GB at 32K.

Where can I download Grug 3b GGUF?

The files are on Hugging Face in the bartowski/ProCreations_grug-3b-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 Grug 3b 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 Grug 3b on 8 GB VRAM?

Yes, the Q4_0 quantization requires about 3.8 GB of VRAM at 8k context, which fits within an 8 GB GPU. The base model without quantization would need significantly more memory.

Does Grug 3b support vision tasks?

No, the model has no vision capabilities. It is designed for text‑only reasoning and tool use.

Other GGUFs of grug-3b

Similar models

Browse all →

More from bartowski

All by bartowski →

More Grug models

All Grug →

Other small models

All small models →