LI

A.X-4.0-Light

LightGGUF by MungertUpdated 2025-09-24
Download on Hugging Face
Parameters7.26B
Context16K
Layers28
Architectureqwen2
Smallest quant1.9 GB
Quick answer

A.X-4.0-Light needs about 5.5 GB of memory at Q4_K_M (a 4.5 GB file) with an 8K context window, so it fits on a 6 GB GPU. This repository has 31 quants, from 1.9 GB (TQ1_0) to 14.5 GB (BF16).

A.X-4.0-Light: 7.26B Qwen2.5 LLM optimized for Korean

A.X-4.0-Light is a 7.26B parameter Qwen2 model trained on Korean data. It supports a 16384 token context window and is released under the apache-2.0 license. The model is available in multiple GGUF quantizations for varied memory budgets.

Good at

  • Korean language understanding
  • Efficient token usage
  • Enterprise deployment

Limits

  • No vision support
  • Requires 8k context for estimates
  • Large memory for higher quants

Best for

  • Korean chatbots
  • Business Korean NLP
  • Cost‑effective inference

Who it’s for. Developers building Korean‑centric applications who need a lightweight yet capable LLM. The model is suitable for on‑premise or edge deployment where memory is a constraint.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
TQ1_01.9 GB 2.9 GB
Fits Get file →
IQ1_S2.2 GB 3.1 GB
Fits Get file →
TQ2_02.2 GB 3.2 GB
Fits Get file →
IQ1_M2.3 GB 3.3 GB
Fits Get file →
IQ2_XXS2.4 GB 3.4 GB
Fits Get file →
IQ2_XS2.6 GB 3.6 GB
Fits Get file →
IQ2_S2.7 GB 3.6 GB
Fits Get file →
IQ2_M2.8 GB 3.8 GB
Fits Get file →
Q2_K_S2.9 GB 3.8 GB
Fits Get file →
Q2_K_M3.0 GB 3.9 GB
Fits Get file →
IQ3_XXS3.1 GB 4.1 GB
Fits Get file →
IQ3_XS3.2 GB 4.2 GB
Fits Get file →
IQ3_M3.5 GB 4.5 GB
Fits Get file →
IQ3_S3.5 GB 4.5 GB
Fits Get file →
Q3_K_S3.6 GB 4.6 GB
Fits Get file →
Q3_K_M3.7 GB 4.7 GB
Fits Get file →
IQ4_XS4.0 GB 4.9 GB
Fits Get file →
Q4_04.1 GB 5.1 GB
Fits Get file →
IQ4_NL4.2 GB 5.2 GB
Fits Get file →
Q4_K_S4.3 GB 5.3 GB
Fits Get file →
Q4_K_M4.5 GB 5.5 GB
Fits Get file →
Q4_14.5 GB 5.5 GB
Fits Get file →
Q5_05.0 GB 6.0 GB
Fits Get file →
Q5_K_S5.2 GB 6.1 GB
Fits Get file →
Q5_K_M5.2 GB 6.2 GB
Fits Get file →
Q5_15.5 GB 6.4 GB
Fits Get file →
Q6_K_M6.0 GB 6.9 GB
Fits Get file →
Q8_07.7 GB 8.7 GB
Too big Get file →
BF1610.6 GB 11.5 GB
Too big Get file →
F1610.6 GB 11.5 GB
Too big Get file →
BF1614.5 GB 15.5 GB
Too big Get file →

For 8 GB VRAM, use TQ1_0 (file size 1.89 GB, memory 2.9 GB) or IQ1_S (2.17 GB, 3.1 GB). With 12 GB VRAM, Q4_0 (4.09 GB, 5.1 GB) or Q4_K_M (4.48 GB, 5.5 GB) are suitable. Higher quants like Q6_K_M (5.96 GB, 6.9 GB) need 16 GB or more.

What GPU do you need for A.X-4.0-Light?

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 Q2_K_M (3.0 GB file, about 3.9 GB in use).
  • 6 GB VRAM: the best quant that fits is Q5_0 (5.0 GB file, about 6.0 GB in use).
  • 8 GB VRAM: the best quant that fits is Q6_K_M (6.0 GB file, about 6.9 GB in use).
  • 12 GB VRAM: the best quant that fits is BF16 (10.6 GB file, about 11.5 GB in use).
  • 16 GB VRAM: the best quant that fits is BF16 (14.5 GB file, about 15.5 GB in use).

Which A.X-4.0-Light quant should you choose?

Q4_K_M is the usual starting point for A.X-4.0-Light: it balances file size and quality well. Here it is a 4.5 GB file.

The Q8_0 file is 7.7 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 TQ1_0 at 1.9 GB, with the quality trade-offs that come with fewer bits. Lower bit counts mean smaller files and more quality loss.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architectureqwen2
Parameters7.26B
Layers28
Trained context16,384 tokens
KV cache (fp16)56 KB per token

Source

Quantized byMungert
FamilyLight
Licenseapache-2.0
Downloads (HF)1,149
Last updated2025-09-24

Frequently asked questions about A.X-4.0-Light

How much VRAM does A.X-4.0-Light need?

About 5.5 GB at Q4_K_M with an 8K context window. That is the 4.5 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 A.X-4.0-Light on an 8 GB GPU?

Yes, with Q6_K_M: the 6.0 GB file needs about 6.9 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of A.X-4.0-Light GGUF?

It depends on the quant: TQ1_0 is 1.9 GB, IQ1_S is 2.2 GB, TQ2_0 is 2.2 GB, IQ1_M is 2.3 GB, IQ2_XXS is 2.4 GB, IQ2_XS is 2.6 GB and 25 more.

What context length does A.X-4.0-Light support?

The GGUF header lists a trained context length of 16,384 tokens (16K). The KV cache grows with the context you use: about 0.5 GB at 8K and 0.9 GB at 16K.

Where can I download A.X-4.0-Light GGUF?

The files are on Hugging Face in the Mungert/A.X-4.0-Light-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 A.X-4.0-Light 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 A.X-4.0-Light on 8 GB VRAM?

Yes, use TQ1_0 or IQ1_S quantizations; they require about 3 GB of VRAM at 8k context.

Does the model support vision tasks?

No, the model has no vision capabilities.

What license applies to this model?

The model is released under the apache-2.0 license.

Similar models

Browse all →

More from Mungert

All by Mungert →

More Light models

All Light →

Other small models

All small models →