Hugging Face's logo

LGAI-EXAONE
/
EXAONE-4.0-32B-GGUF

Text Generation
Transformers
GGUF
English
Korean
Spanish
lg-ai
exaone
exaone-4.0
conversational
Model card Files
xet
Community
7
EXAONE-4.0-32B-GGUF / README.md
LG-AI-EXAONE's picture
LG-AI-EXAONE
Update README.md
b390c77
preview code
|
raw
history blame contribute delete
35.9 kB
 ---
base_model: LGAI-EXAONE/EXAONE-4.0-32B
base_model_relation: quantized
license: other
license_name: exaone
license_link: LICENSE
language:
- en
- ko
- es
tags:
- lg-ai
- exaone
- exaone-4.0
pipeline_tag: text-generation
library_name: transformers
---
<p align="center">
<img src="assets/EXAONE_Symbol+BI_3d.png", width="300", style="margin: 40 auto;">
🎉 License Updated! We are pleased to announce our more flexible licensing terms 🤗
<br>✈️ Try on <a href="https://friendli.ai/suite/~/serverless-endpoints/LGAI-EXAONE/EXAONE-4.0-32B/overview">FriendliAI</a> (licensed under commercial purposes)
<br><br><i>📢 EXAONE 4.0 is officially supported by llama.cpp! Please check the guide <a href="#quickstart">below</a></i>
<br>
# EXAONE-4.0-32B-GGUF
## Introduction
We introduce **EXAONE 4.0**, which integrates a **Non-reasoning mode** and **Reasoning mode** to achieve both the excellent usability of [EXAONE 3.5](https://github.com/LG-AI-EXAONE/EXAONE-3.5) and the advanced reasoning abilities of [EXAONE Deep](https://github.com/LG-AI-EXAONE/EXAONE-Deep). To pave the way for the agentic AI era, EXAONE 4.0 incorporates essential features such as agentic tool use, and its multilingual capabilities are extended
to support Spanish in addition to English and Korean.
The EXAONE 4.0 model series consists of two sizes: a mid-size **32B** model optimized for high performance, and a small-size **1.2B** model designed for on-device applications.
In the EXAONE 4.0 architecture, we apply new architectural changes compared to previous EXAONE models as below:
1. **Hybrid Attention**: For the 32B model, we adopt hybrid attention scheme, which combines *Local attention (sliding window attention)* with *Global attention (full attention)* in a 3:1 ratio. We do not use RoPE (Rotary Positional Embedding) for global attention for better global context understanding.
2. **QK-Reorder-Norm**: We reorder the LayerNorm position from the traditional Pre-LN scheme by applying LayerNorm directly to the attention and MLP outputs, and we add RMS normalization right after the Q and K projection. It helps yield better performance on downstream tasks despite consuming more computation.
For more details, please refer to our [technical report](https://arxiv.org/abs/2507.11407), [HuggingFace paper](https://huggingface.co/papers/2507.11407), [blog](https://www.lgresearch.ai/blog/view?seq=576), and [GitHub](https://github.com/LG-AI-EXAONE/EXAONE-4.0).
### Model Configuration
- Number of Parameters (without embeddings): 30.95B
- Number of Layers: 64
- Number of Attention Heads: GQA with 40-heads and 8-KV heads
- Vocab Size: 102,400
- Context Length: 131,072 tokens
- Quantization: `Q8_0`, `Q6_K`, `Q5_K_M`, `Q4_K_M`, `IQ4_XS` in GGUF format (also includes `BF16` weights)
## Quickstart
### llama.cpp
You can run EXAONE models locally using llama.cpp by following these steps:
1. Install the latest version of llama.cpp (version >= `b5932`). Please check the official [installation guide](https://github.com/ggml-org/llama.cpp?tab=readme-ov-file#quick-start) from llama.cpp.
2. Download the EXAONE 4.0 model weights in GGUF format.
```bash
huggingface-cli download LGAI-EXAONE/EXAONE-4.0-32B-GGUF \
--include "EXAONE-4.0-32B-Q4_K_M.gguf" \
--local-dir .
```
When you use GGUF model split into multiple files, you should merge them into a single file before running the model.
1. First, download the GGUF model weights.
```bash
huggingface-cli download LGAI-EXAONE/EXAONE-4.0-32B-GGUF \
--include "EXAONE-4.0-32B-BF16*.gguf" \
--local-dir .
```
2. Merge the split files into a single file.
```bash
llama-gguf-split --merge \
./EXAONE-4.0-32B-BF16-00001-of-00002.gguf \
./EXAONE-4.0-32B-BF16.gguf
```
<details>
<summary>Generation with `llama-cli`</summary>
3. Apply chat template using transformers.
> This process is necessary to avoid issues with current EXAONE modeling code in `llama.cpp`. This is work in progress at our [PR](https://github.com/ggml-org/llama.cpp/pull/14630). We will update this once these issues are solved.
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
model_name = "LGAI-EXAONE/EXAONE-4.0-32B"
tokenizer = AutoTokenizer.from_pretrained(model_name)
messages = [
{"role": "user", "content": "Let's work together on local system!"}
]
input_text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True,
)
print(repr(input_text))
with open("inputs.txt", "w") as f:
f.write(input_text)
```
4. Generate result with greedy decoding.
```bash
llama-cli -m EXAONE-4.0-32B-Q4_K_M.gguf \
-fa -ngl 65 \
--temp 0.0 --top-k 1 \
-f inputs.txt -no-cnv
```
</details>
<details>
<summary>OpenAI compatible server with `llama-server`</summary>
3. Run llama-server with EXAONE 4.0 Jinja template. You can find the [chat template file](https://huggingface.co/LGAI-EXAONE/EXAONE-4.0-32B-GGUF/blob/main/chat_template.jinja) in this repository.
```bash
llama-server -m EXAONE-4.0-32B-Q4_K_M.gguf \
-c 131072 -fa -ngl 65 \
--temp 0.6 --top-p 0.95 \
--jinja --chat-template-file chat_template.jinja \
--host 0.0.0.0 --port 8820 \
-a EXAONE-4.0-32B-Q4_K_M
```
4. Use OpenAI chat completion to test the GGUF model.
```bash
curl -X POST http://localhost:8820/v1/chat/completions \
-H "Content-Type: application/json" \
-d '{
"model": "EXAONE-4.0-32B-Q4_K_M",
"messages": [
{"role": "user", "content": "Let'\''s work together on server!"}
],
"max_tokens": 1024,
"temperature": 0.6,
"top_p": 0.95,
"chat_template_kwargs": {"enable_thinking": false}
}'
```
</details>
## Performance
The following tables show the evaluation results of each model, with reasoning and non-reasoning mode. The evaluation details can be found in the [technical report](https://arxiv.org/abs/2507.11407).
- ✅ denotes the model has a hybrid reasoning capability, evaluated by selecting reasoning / non-reasoning on the purpose.
- To assess Korean **practical** and **professional** knowledge, we adopt both the [KMMLU-Redux](https://huggingface.co/datasets/LGAI-EXAONE/KMMLU-Redux) and [KMMLU-Pro](https://huggingface.co/datasets/LGAI-EXAONE/KMMLU-Pro) benchmarks. Both datasets are publicly released!
- The evaluation results are based on the original model, not quantized model.
### 32B Reasoning Mode
<table>
<tr>
<th> </th>
<th>EXAONE 4.0 32B </th>
<th>Phi 4 reasoning-plus</th>
<th>Magistral Small-2506</th>
<th>Qwen 3 32B </th>
<th>Qwen 3 235B </th>
<th>DeepSeek R1-0528</th>
</tr>
<tr>
<td align="center">Model Size</td>
<td align="center">32.0B</td>
<td align="center">14.7B</td>
<td align="center">23.6B</td>
<td align="center">32.8B</td>
<td align="center">235B</td>
<td align="center">671B</td>
</tr>
<tr>
<td align="center">Hybrid Reasoning</td>
<td align="center">✅</td>
<td align="center"> </td>
<td align="center"> </td>
<td align="center">✅</td>
<td align="center">✅</td>
<td align="center"> </td>
</tr>
<tr>
<td align="center" colspan='7'><i>World Knowledge</i></td>
</tr>
<tr>
<td >MMLU-Redux</td>
<td align="center">92.3</td>
<td align="center">90.8</td>
<td align="center">86.8</td>
<td align="center">90.9</td>
<td align="center">92.7</td>
<td align="center">93.4</td>
</tr>
<tr>
<td >MMLU-Pro</td>
<td align="center">81.8</td>
<td align="center">76.0</td>
<td align="center">73.4</td>
<td align="center">80.0</td>
<td align="center">83.0</td>
<td align="center">85.0</td>
</tr>
<tr>
<td >GPQA-Diamond</td>
<td align="center">75.4</td>
<td align="center">68.9</td>
<td align="center">68.2</td>
<td align="center">68.4</td>
<td align="center">71.1</td>
<td align="center">81.0</td>
</tr>
<tr>
<td align="center" colspan='7'><i>Math/Coding</i></td>
</tr>
<tr>
<td >AIME 2025</td>
<td align="center">85.3</td>
<td align="center">78.0</td>
<td align="center">62.8</td>
<td align="center">72.9</td>
<td align="center">81.5</td>
<td align="center">87.5</td>
</tr>
<tr>
<td >HMMT Feb 2025</td>
<td align="center">72.9</td>
<td align="center">53.6</td>
<td align="center">43.5</td>
<td align="center">50.4</td>
<td align="center">62.5</td>
<td align="center">79.4</td>
</tr>
<tr>
<td >LiveCodeBench v5</td>
<td align="center">72.6</td>
<td align="center">51.7</td>
<td align="center">55.8</td>
<td align="center">65.7</td>
<td align="center">70.7</td>
<td align="center">75.2</td>
</tr>
<tr>
<td >LiveCodeBench v6</td>
<td align="center">66.7</td>
<td align="center">47.1</td>
<td align="center">47.4</td>
<td align="center">60.1</td>
<td align="center">58.9</td>
<td align="center">70.3</td>
</tr>
<tr>
<td align="center" colspan='7'><i>Instruction Following</i></td>
</tr>
<tr>
<td >IFEval</td>
<td align="center">83.7</td>
<td align="center">84.9</td>
<td align="center">37.9</td>
<td align="center">85.0</td>
<td align="center">83.4</td>
<td align="center">80.8</td>
</tr>
<tr>
<td >Multi-IF (EN)</td>
<td align="center">73.5</td>
<td align="center">56.1</td>
<td align="center">27.4</td>
<td align="center">73.4</td>
<td align="center">73.4</td>
<td align="center">72.0</td>
</tr>
<tr>
<td align="center" colspan='7'><i>Agentic Tool Use</i></td>
</tr>
<tr>
<td >BFCL-v3</td>
<td align="center">63.9</td>
<td align="center">N/A</td>
<td align="center">40.4</td>
<td align="center">70.3</td>
<td align="center">70.8</td>
<td align="center">64.7</td>
</tr>
<tr>
<td >Tau-Bench (Airline)</td>
<td align="center">51.5</td>
<td align="center">N/A</td>
<td align="center">38.5</td>
<td align="center">34.5</td>
<td align="center">37.5</td>
<td align="center">53.5</td>
</tr>
<tr>
<td >Tau-Bench (Retail)</td>
<td align="center">62.8</td>
<td align="center">N/A</td>
<td align="center">10.2</td>
<td align="center">55.2</td>
<td align="center">58.3</td>
<td align="center">63.9</td>
</tr>
<tr>
<td align="center" colspan='7'><i>Multilinguality</i></td>
</tr>
<tr>
<td >KMMLU-Pro</td>
<td align="center">67.7</td>
<td align="center">55.8</td>
<td align="center">51.5</td>
<td align="center">61.4</td>
<td align="center">68.1</td>
<td align="center">71.7</td>
</tr>
<tr>
<td >KMMLU-Redux</td>
<td align="center">72.7</td>
<td align="center">62.7</td>
<td align="center">54.6</td>
<td align="center">67.5</td>
<td align="center">74.5</td>
<td align="center">77.0</td>
</tr>
<tr>
<td >KSM</td>
<td align="center">87.6</td>
<td align="center">79.8</td>
<td align="center">71.9</td>
<td align="center">82.8</td>
<td align="center">86.2</td>
<td align="center">86.7</td>
</tr>
<tr>
<td >MMMLU (ES)</td>
<td align="center">85.6</td>
<td align="center">84.3</td>
<td align="center">68.9</td>
<td align="center">82.8</td>
<td align="center">86.7</td>
<td align="center">88.2</td>
</tr>
<tr>
<td >MATH500 (ES)</td>
<td align="center">95.8</td>
<td align="center">94.2</td>
<td align="center">83.5</td>
<td align="center">94.3</td>
<td align="center">95.1</td>
<td align="center">96.0</td>
</tr>
</table>
### 32B Non-Reasoning Mode
<table>
<tr>
<th> </th>
<th>EXAONE 4.0 32B </th>
<th>Phi 4</th>
<th>Mistral-Small-2506</th>
<th>Gemma3 27B</th>
<th>Qwen3 32B </th>
<th>Qwen3 235B </th>
<th>Llama-4-Maverick</th>
<th>DeepSeek V3-0324</th>
</tr>
<tr>
<td align="center">Model Size</td>
<td align="center">32.0B</td>
<td align="center">14.7B</td>
<td align="center">24.0B</td>
<td align="center">27.4B</td>
<td align="center">32.8B</td>
<td align="center">235B</td>
<td align="center">402B</td>
<td align="center">671B</td>
</tr>
<tr>
<td align="center">Hybrid Reasoning</td>
<td align="center">✅</td>
<td align="center"> </td>
<td align="center"> </td>
<td align="center"> </td>
<td align="center">✅</td>
<td align="center">✅</td>
<td align="center"> </td>
<td align="center"> </td>
</tr>
<tr>
<td align="center" colspan='9'><i>World Knowledge</i></td>
</tr>
<tr>
<td >MMLU-Redux</td>
<td align="center">89.8</td>
<td align="center">88.3</td>
<td align="center">85.9</td>
<td align="center">85.0</td>
<td align="center">85.7</td>
<td align="center">89.2</td>
<td align="center">92.3</td>
<td align="center">92.3</td>
</tr>
<tr>
<td >MMLU-Pro</td>
<td align="center">77.6</td>
<td align="center">70.4</td>
<td align="center">69.1</td>
<td align="center">67.5</td>
<td align="center">74.4</td>
<td align="center">77.4</td>
<td align="center">80.5</td>
<td align="center">81.2</td>
</tr>
<tr>
<td >GPQA-Diamond</td>
<td align="center">63.7</td>
<td align="center">56.1</td>
<td align="center">46.1</td>
<td align="center">42.4</td>
<td align="center">54.6</td>
<td align="center">62.9</td>
<td align="center">69.8</td>
<td align="center">68.4</td>
</tr>
<tr>
<td align="center" colspan='9'><i>Math/Coding</i></td>
</tr>
<tr>
<td >AIME 2025</td>
<td align="center">35.9</td>
<td align="center">17.8</td>
<td align="center">30.2</td>
<td align="center">23.8</td>
<td align="center">20.2</td>
<td align="center">24.7</td>
<td align="center">18.0</td>
<td align="center">50.0</td>
</tr>
<tr>
<td >HMMT Feb 2025</td>
<td align="center">21.8</td>
<td align="center">4.0</td>
<td align="center">16.9</td>
<td align="center">10.3</td>
<td align="center">9.8</td>
<td align="center">11.9</td>
<td align="center">7.3</td>
<td align="center">29.2</td>
</tr>
<tr>
<td >LiveCodeBench v5</td>
<td align="center">43.3</td>
<td align="center">24.6</td>
<td align="center">25.8</td>
<td align="center">27.5</td>
<td align="center">31.3</td>
<td align="center">35.3</td>
<td align="center">43.4</td>
<td align="center">46.7</td>
</tr>
<tr>
<td >LiveCodeBench v6</td>
<td align="center">43.1</td>
<td align="center">27.4</td>
<td align="center">26.9</td>
<td align="center">29.7</td>
<td align="center">28.0</td>
<td align="center">31.4</td>
<td align="center">32.7</td>
<td align="center">44.0</td>
</tr>
<tr>
<td align="center" colspan='9'><i>Instruction Following</i></td>
</tr>
<tr>
<td >IFEval</td>
<td align="center">84.8</td>
<td align="center">63.0</td>
<td align="center">77.8</td>
<td align="center">82.6</td>
<td align="center">83.2</td>
<td align="center">83.2</td>
<td align="center">85.4</td>
<td align="center">81.2</td>
</tr>
<tr>
<td >Multi-IF (EN)</td>
<td align="center">71.6</td>
<td align="center">47.7</td>
<td align="center">63.2</td>
<td align="center">72.1</td>
<td align="center">71.9</td>
<td align="center">72.5</td>
<td align="center">77.9</td>
<td align="center">68.3</td>
</tr>
<tr>
<td align="center" colspan='9'><i>Long Context</i></td>
</tr>
<tr>
<td >HELMET</td>
<td align="center">58.3</td>
<td align="center">N/A</td>
<td align="center">61.9</td>
<td align="center">58.3</td>
<td align="center">54.5</td>
<td align="center">63.3</td>
<td align="center">13.7</td>
<td align="center">N/A</td>
</tr>
<tr>
<td >RULER</td>
<td align="center">88.2</td>
<td align="center">N/A</td>
<td align="center">71.8</td>
<td align="center">66.0</td>
<td align="center">85.6</td>
<td align="center">90.6</td>
<td align="center">2.9</td>
<td align="center">N/A</td>
</tr>
<tr>
<td >LongBench v1</td>
<td align="center">48.1</td>
<td align="center">N/A</td>
<td align="center">51.5</td>
<td align="center">51.5</td>
<td align="center">44.2</td>
<td align="center">45.3</td>
<td align="center">34.7</td>
<td align="center">N/A</td>
</tr>
<tr>
<td align="center" colspan='9'><i>Agentic Tool Use</i></td>
</tr>
<tr>
<td >BFCL-v3</td>
<td align="center">65.2</td>
<td align="center">N/A</td>
<td align="center">57.7</td>
<td align="center">N/A</td>
<td align="center">63.0</td>
<td align="center">68.0</td>
<td align="center">52.9</td>
<td align="center">63.8</td>
</tr>
<tr>
<td >Tau-Bench (Airline)</td>
<td align="center">25.5</td>
<td align="center">N/A</td>
<td align="center">36.1</td>
<td align="center">N/A</td>
<td align="center">16.0</td>
<td align="center">27.0</td>
<td align="center">38.0</td>
<td align="center">40.5</td>
</tr>
<tr>
<td >Tau-Bench (Retail)</td>
<td align="center">55.9</td>
<td align="center">N/A</td>
<td align="center">35.5</td>
<td align="center">N/A</td>
<td align="center">47.6</td>
<td align="center">56.5</td>
<td align="center">6.5</td>
<td align="center">68.5</td>
</tr>
<tr>
<td align="center" colspan='9'><i>Multilinguality</i></td>
</tr>
<tr>
<td >KMMLU-Pro</td>
<td align="center">60.0</td>
<td align="center">44.8</td>
<td align="center">51.0</td>
<td align="center">50.7</td>
<td align="center">58.3</td>
<td align="center">64.4</td>
<td align="center">68.8</td>
<td align="center">67.3</td>
</tr>
<tr>
<td >KMMLU-Redux</td>
<td align="center">64.8</td>
<td align="center">50.1</td>
<td align="center">53.6</td>
<td align="center">53.3</td>
<td align="center">64.4</td>
<td align="center">71.7</td>
<td align="center">76.9</td>
<td align="center">72.2</td>
</tr>
<tr>
<td >KSM</td>
<td align="center">59.8</td>
<td align="center">29.1</td>
<td align="center">35.5</td>
<td align="center">36.1</td>
<td align="center">41.3</td>
<td align="center">46.6</td>
<td align="center">40.6</td>
<td align="center">63.5</td>
</tr>
<tr>
<td >Ko-LongBench</td>
<td align="center">76.9</td>
<td align="center">N/A</td>
<td align="center">55.4</td>
<td align="center">72.0</td>
<td align="center">73.9</td>
<td align="center">74.6</td>
<td align="center">65.6</td>
<td align="center">N/A</td>
</tr>
<tr>
<td >MMMLU (ES)</td>
<td align="center">80.6</td>
<td align="center">81.2</td>
<td align="center">78.4</td>
<td align="center">78.7</td>
<td align="center">82.1</td>
<td align="center">83.7</td>
<td align="center">86.9</td>
<td align="center">86.7</td>
</tr>
<tr>
<td >MATH500 (ES)</td>
<td align="center">87.3</td>
<td align="center">78.2</td>
<td align="center">83.4</td>
<td align="center">86.8</td>
<td align="center">84.7</td>
<td align="center">87.2</td>
<td align="center">78.7</td>
<td align="center">89.2</td>
</tr>
<tr>
<td >WMT24++ (ES)</td>
<td align="center">90.7</td>
<td align="center">89.3</td>
<td align="center">92.2</td>
<td align="center">93.1</td>
<td align="center">91.4</td>
<td align="center">92.9</td>
<td align="center">92.7</td>
<td align="center">94.3 </td>
</tr>
</table>
### 1.2B Reasoning Mode
<table>
<tr>
<th> </th>
<th>EXAONE 4.0 1.2B </th>
<th>EXAONE Deep 2.4B</th>
<th>Qwen 3 0.6B </th>
<th>Qwen 3 1.7B </th>
<th>SmolLM 3 3B </th>
</tr>
<tr>
<td align="center">Model Size</td>
<td align="center">1.28B</td>
<td align="center">2.41B</td>
<td align="center">596M</td>
<td align="center">1.72B</td>
<td align="center">3.08B</td>
</tr>
<tr>
<td align="center">Hybrid Reasoning</td>
<td align="center">✅</td>
<td align="center"> </td>
<td align="center">✅</td>
<td align="center">✅</td>
<td align="center">✅</td>
</tr>
<tr>
<td align="center" colspan='6'><i>World Knowledge</i></td>
</tr>
<tr>
<td >MMLU-Redux</td>
<td align="center">71.5</td>
<td align="center">68.9</td>
<td align="center">55.6</td>
<td align="center">73.9</td>
<td align="center">74.8</td>
</tr>
<tr>
<td >MMLU-Pro</td>
<td align="center">59.3</td>
<td align="center">56.4</td>
<td align="center">38.3</td>
<td align="center">57.7</td>
<td align="center">57.8</td>
</tr>
<tr>
<td >GPQA-Diamond</td>
<td align="center">52.0</td>
<td align="center">54.3</td>
<td align="center">27.9</td>
<td align="center">40.1</td>
<td align="center">41.7</td>
</tr>
<tr>
<td align="center" colspan='6'><i>Math/Coding</i></td>
</tr>
<tr>
<td >AIME 2025</td>
<td align="center">45.2</td>
<td align="center">47.9</td>
<td align="center">15.1</td>
<td align="center">36.8</td>
<td align="center">36.7</td>
</tr>
<tr>
<td >HMMT Feb 2025</td>
<td align="center">34.0</td>
<td align="center">27.3</td>
<td align="center">7.0</td>
<td align="center">21.8</td>
<td align="center">26.0</td>
</tr>
<tr>
<td >LiveCodeBench v5</td>
<td align="center">44.6</td>
<td align="center">47.2</td>
<td align="center">12.3</td>
<td align="center">33.2</td>
<td align="center">27.6</td>
</tr>
<tr>
<td >LiveCodeBench v6</td>
<td align="center">45.3</td>
<td align="center">43.1</td>
<td align="center">16.4</td>
<td align="center">29.9</td>
<td align="center">29.1</td>
</tr>
<tr>
<td align="center" colspan='6'><i>Instruction Following</i></td>
</tr>
<tr>
<td >IFEval</td>
<td align="center">67.8</td>
<td align="center">71.0</td>
<td align="center">59.2</td>
<td align="center">72.5</td>
<td align="center">71.2</td>
</tr>
<tr>
<td >Multi-IF (EN)</td>
<td align="center">53.9</td>
<td align="center">54.5</td>
<td align="center">37.5</td>
<td align="center">53.5</td>
<td align="center">47.5</td>
</tr>
<tr>
<td align="center" colspan='6'><i>Agentic Tool Use</i></td>
</tr>
<tr>
<td >BFCL-v3</td>
<td align="center">52.9</td>
<td align="center">N/A</td>
<td align="center">46.4</td>
<td align="center">56.6</td>
<td align="center">37.1</td>
</tr>
<tr>
<td >Tau-Bench (Airline)</td>
<td align="center">20.5</td>
<td align="center">N/A</td>
<td align="center">22.0</td>
<td align="center">31.0</td>
<td align="center">37.0</td>
</tr>
<tr>
<td >Tau-Bench (Retail)</td>
<td align="center">28.1</td>
<td align="center">N/A</td>
<td align="center">3.3</td>
<td align="center">6.5</td>
<td align="center">5.4</td>
</tr>
<tr>
<td align="center" colspan='6'><i>Multilinguality</i></td>
</tr>
<tr>
<td >KMMLU-Pro</td>
<td align="center">42.7</td>
<td align="center">24.6</td>
<td align="center">21.6</td>
<td align="center">38.3</td>
<td align="center">30.5</td>
</tr>
<tr>
<td >KMMLU-Redux</td>
<td align="center">46.9</td>
<td align="center">25.0</td>
<td align="center">24.5</td>
<td align="center">38.0</td>
<td align="center">33.7</td>
</tr>
<tr>
<td >KSM</td>
<td align="center">60.6</td>
<td align="center">60.9</td>
<td align="center">22.8</td>
<td align="center">52.9</td>
<td align="center">49.7</td>
</tr>
<tr>
<td >MMMLU (ES)</td>
<td align="center">62.4</td>
<td align="center">51.4</td>
<td align="center">48.8</td>
<td align="center">64.5</td>
<td align="center">64.7</td>
</tr>
<tr>
<td >MATH500 (ES)</td>
<td align="center">88.8</td>
<td align="center">84.5</td>
<td align="center">70.6</td>
<td align="center">87.9</td>
<td align="center">87.5 </td>
</tr>
</table>
### 1.2B Non-Reasoning Mode
<table>
<tr>
<th> </th>
<th>EXAONE 4.0 1.2B </th>
<th>Qwen 3 0.6B </th>
<th>Gemma 3 1B</th>
<th>Qwen 3 1.7B </th>
<th>SmolLM 3 3B </th>
</tr>
<tr>
<td align="center">Model Size</td>
<td align="center">1.28B</td>
<td align="center">596M</td>
<td align="center">1.00B</td>
<td align="center">1.72B</td>
<td align="center">3.08B</td>
</tr>
<tr>
<td align="center">Hybrid Reasoning</td>
<td align="center">✅</td>
<td align="center">✅</td>
<td align="center"> </td>
<td align="center">✅</td>
<td align="center">✅</td>
</tr>
<tr>
<td align="center" colspan='6'><i>World Knowledge</i></td>
</tr>
<tr>
<td >MMLU-Redux</td>
<td align="center">66.9</td>
<td align="center">44.6</td>
<td align="center">40.9</td>
<td align="center">63.4</td>
<td align="center">65.0</td>
</tr>
<tr>
<td >MMLU-Pro</td>
<td align="center">52.0</td>
<td align="center">26.6</td>
<td align="center">14.7</td>
<td align="center">43.7</td>
<td align="center">43.6</td>
</tr>
<tr>
<td >GPQA-Diamond</td>
<td align="center">40.1</td>
<td align="center">22.9</td>
<td align="center">19.2</td>
<td align="center">28.6</td>
<td align="center">35.7</td>
</tr>
<tr>
<td align="center" colspan='6'><i>Math/Coding</i></td>
</tr>
<tr>
<td >AIME 2025</td>
<td align="center">23.5</td>
<td align="center">2.6</td>
<td align="center">2.1</td>
<td align="center">9.8</td>
<td align="center">9.3</td>
</tr>
<tr>
<td >HMMT Feb 2025</td>
<td align="center">13.0</td>
<td align="center">1.0</td>
<td align="center">1.5</td>
<td align="center">5.1</td>
<td align="center">4.7</td>
</tr>
<tr>
<td >LiveCodeBench v5</td>
<td align="center">26.4</td>
<td align="center">3.6</td>
<td align="center">1.8</td>
<td align="center">11.6</td>
<td align="center">11.4</td>
</tr>
<tr>
<td >LiveCodeBench v6</td>
<td align="center">30.1</td>
<td align="center">6.9</td>
<td align="center">2.3</td>
<td align="center">16.6</td>
<td align="center">20.6</td>
</tr>
<tr>
<td align="center" colspan='6'><i>Instruction Following</i></td>
</tr>
<tr>
<td >IFEval</td>
<td align="center">74.7</td>
<td align="center">54.5</td>
<td align="center">80.2</td>
<td align="center">68.2</td>
<td align="center">76.7</td>
</tr>
<tr>
<td >Multi-IF (EN)</td>
<td align="center">62.1</td>
<td align="center">37.5</td>
<td align="center">32.5</td>
<td align="center">51.0</td>
<td align="center">51.9</td>
</tr>
<tr>
<td align="center" colspan='6'><i>Long Context</i></td>
</tr>
<tr>
<td >HELMET</td>
<td align="center">41.2</td>
<td align="center">21.1</td>
<td align="center">N/A</td>
<td align="center">33.8</td>
<td align="center">38.6</td>
</tr>
<tr>
<td >RULER</td>
<td align="center">77.4</td>
<td align="center">55.1</td>
<td align="center">N/A</td>
<td align="center">65.9</td>
<td align="center">66.3</td>
</tr>
<tr>
<td >LongBench v1</td>
<td align="center">36.9</td>
<td align="center">32.4</td>
<td align="center">N/A</td>
<td align="center">41.9</td>
<td align="center">39.9</td>
</tr>
<tr>
<td align="center" colspan='6'><i>Agentic Tool Use</i></td>
</tr>
<tr>
<td >BFCL-v3</td>
<td align="center">55.7</td>
<td align="center">44.1</td>
<td align="center">N/A</td>
<td align="center">52.2</td>
<td align="center">47.3</td>
</tr>
<tr>
<td >Tau-Bench (Airline)</td>
<td align="center">10.0</td>
<td align="center">31.5</td>
<td align="center">N/A</td>
<td align="center">13.5</td>
<td align="center">38.0</td>
</tr>
<tr>
<td >Tau-Bench (Retail)</td>
<td align="center">21.7</td>
<td align="center">5.7</td>
<td align="center">N/A</td>
<td align="center">4.6</td>
<td align="center">6.7</td>
</tr>
<tr>
<td align="center" colspan='6'><i>Multilinguality</i></td>
</tr>
<tr>
<td >KMMLU-Pro</td>
<td align="center">37.5</td>
<td align="center">24.6</td>
<td align="center">9.7</td>
<td align="center">29.5</td>
<td align="center">27.6</td>
</tr>
<tr>
<td >KMMLU-Redux</td>
<td align="center">40.4</td>
<td align="center">22.8</td>
<td align="center">19.4</td>
<td align="center">29.8</td>
<td align="center">26.4</td>
</tr>
<tr>
<td >KSM</td>
<td align="center">26.3</td>
<td align="center">0.1</td>
<td align="center">22.8</td>
<td align="center">16.3</td>
<td align="center">16.1</td>
</tr>
<tr>
<td >Ko-LongBench</td>
<td align="center">69.8</td>
<td align="center">16.4</td>
<td align="center">N/A</td>
<td align="center">57.1</td>
<td align="center">15.7</td>
</tr>
<tr>
<td >MMMLU (ES)</td>
<td align="center">54.6</td>
<td align="center">39.5</td>
<td align="center">35.9</td>
<td align="center">54.3</td>
<td align="center">55.1</td>
</tr>
<tr>
<td >MATH500 (ES)</td>
<td align="center">71.2</td>
<td align="center">38.5</td>
<td align="center">41.2</td>
<td align="center">66.0</td>
<td align="center">62.4</td>
</tr>
<tr>
<td >WMT24++ (ES)</td>
<td align="center">65.9</td>
<td align="center">58.2</td>
<td align="center">76.9</td>
<td align="center">76.7</td>
<td align="center">84.0 </td>
</tr>
</table>
## Usage Guideline
> [!IMPORTANT]
> To achieve the expected performance, we recommend using the following configurations:
>
> - For non-reasoning mode, we recommend using a lower temperature value such as `temperature<0.6` for better performance.
> - For reasoning mode (using `<think>` block), we recommend using `temperature=0.6` and `top_p=0.95`.
> - If you suffer from the model degeneration, we recommend using `presence_penalty=1.5`.
> - For Korean general conversation with 1.2B model, we suggest to use `temperature=0.1` to avoid code switching.
## Limitation
The EXAONE language model has certain limitations and may occasionally generate inappropriate responses. The language model generates responses based on the output probability of tokens, and it is determined during learning from training data. While we have made every effort to exclude personal, harmful, and biased information from the training data, some problematic content may still be included, potentially leading to undesirable responses. Please note that the text generated by EXAONE language model does not reflect the views of LG AI Research.
- Inappropriate answers may be generated, which contain personal, harmful or other inappropriate information.
- Biased responses may be generated, which are associated with age, gender, race, and so on.
- The generated responses rely heavily on statistics from the training data, which can result in the generation of
semantically or syntactically incorrect sentences.
- Since the model does not reflect the latest information, the responses may be false or contradictory.
LG AI Research strives to reduce potential risks that may arise from EXAONE language models. Users are not allowed
to engage in any malicious activities (e.g., keying in illegal information) that may induce the creation of inappropriate
outputs violating LG AI's ethical principles when using EXAONE language models.
## License
The model is licensed under [EXAONE AI Model License Agreement 1.2 - NC](./LICENSE)
> [!NOTE]
> The main difference from the older version is as below:
> - We removed **the claim of model output ownership** from the license.
> - We restrict the model use **against the development of models that compete with EXAONE**.
> - We allow the model to be used for **educational purposes**, not just research.
## Citation
```
@article{exaone-4.0,
title={EXAONE 4.0: Unified Large Language Models Integrating Non-reasoning and Reasoning Modes},
author={{LG AI Research}},
journal={arXiv preprint arXiv:2507.11407},
year={2025}
}
```
## Contact
LG AI Research Technical Support: contact[email protected]