---
title: "LG's EXAONE 3.5 Hits 1 Million Monthly Downloads Running on 5GB"
subtitle: "LG AI Research's 4-bit quantized 7.8B bilingual model shrinks a strong Korean-English instruction follower into roughly 5GB of weights with a 32K context window."
url: "https://alphasignal.ai/news/lg-s-exaone-3-5-hits-1-million-monthly-downloads-running-on-5gb"
author: "AlphaSignal Newsroom"
author_url: "https://alphasignal.ai/editorial-team"
company: "AlphaSignal"
company_url: "https://alphasignal.ai/company/alphasignal"
published: "2026-09-27T16:00:57.848Z"
updated: "2026-09-27T18:08:28.634Z"
type: "Model"
topics:
  - "Llms"
  - "Gpus"
subtopics:
  - "Small Models"
  - "Long Context"
  - "Quantization"
upvotes: 3462
source_url: "https://huggingface.co/LGAI-EXAONE/EXAONE-3.5-7.8B-Instruct-AWQ"
comments: 0
truncated: true
---

# LG's EXAONE 3.5 Hits 1 Million Monthly Downloads Running on 5GB

> LG AI Research's 4-bit quantized 7.8B bilingual model shrinks a strong Korean-English instruction follower into roughly 5GB of weights with a 32K context window.

# LG’s 4-bit EXAONE 3.5 approaches one million monthly Hugging Face downloads

At publication time, LG AI Research’s AWQ-quantized EXAONE 3.5 7.8B Instruct checkpoint was approaching one million monthly downloads on its [AWQ model page](https://huggingface.co/LGAI-EXAONE/EXAONE-3.5-7.8B-Instruct-AWQ). Hugging Face counts qualifying requests, so the figure does not represent one million unique users or deployments. It does indicate sustained interest in a compact Korean-English model designed for local inference.

The wider EXAONE 3.5 family includes 2.4B, 7.8B, and 32B instruction-tuned models. This 7.8B release uses activation-aware weight quantization, or AWQ, to reduce GPU memory requirements while retaining the model’s configured 32,768-token context window.

## A compact checkpoint with a large window

The checkpoint contains a decoder-only Transformer with the following specifications:

* 6.98 billion non-embedding parameters across 32 layers
* Grouped-query attention with 32 query heads and 8 key/value heads
* A 102,400-token byte-level BPE vocabulary designed around Korean and English
* A maximum sequence length of 32,768 tokens, including input and generated tokens
* W4A16 group-wise quantization with groups of 128 weights

W4A16 stores quantized weights at 4-bit precision while keeping activations at 16-bit precision. Groups of 128 weights share quantization metadata. The resulting files occupy roughly 5 GB, although actual GPU memory use also includes the key-value cache, activations, temporary buffers, and backend overhead.

Quantization leaves the architecture and configured context length unchanged. A full 32K sequence at batch size one can add roughly 4 GiB for a 16-bit key-value cache before other allocations, so loading the weights in about 5 GB does not mean 32K inference will run within that same memory budget.

EXAONE 3.5 retains the broad architecture of EXAONE 3.0 7.8B. According to LG’s [technical report](https://arxiv.org/html/2412.04862), long-context fine-tuning extended the maximum sequence length from 4,096 to 32,768 tokens. The model also uses a rotary-position-embedding theta of 1,000,000 to support the larger window.

_Preview only. Full text requires an AlphaSignal Pro account._

## Sources

- [EXAONE-3.5-7.8B-Instruct-AWQ model card](https://huggingface.co/LGAI-EXAONE/EXAONE-3.5-7.8B-Instruct-AWQ)
- [EXAONE 3.5: Series of Large Language Models for Real-world Use Cases](https://arxiv.org/html/2412.04862)
- [EXAONE 3.5 GitHub repository](https://github.com/LG-AI-EXAONE/EXAONE-3.5)
- [EXAONE 3.5 Hugging Face collection](https://huggingface.co/collections/LGAI-EXAONE/exaone-35)