---
pipeline_tag: image-text-to-text
language:
- multilingual
library_name: transformers
tags:
- llm
- image-text-to-text
- neuralnode
- multilingual
- horus
- tokenai
---

<div align="center">
  <h1><b style="font-size: 1.5em;">Horus Hiero 4B Mini</b></h1>
</div>

<p align="center">
  <img src="media/Horus%20Hiero.png" alt="Horus Hiero 4B Mini" width="600"/>
</p>

<p align="center">
  <b>Horus Hiero 4B Mini</b> is an advanced language model developed by <b>TokenAI</b>, an Egyptian AI startup founded by Assem Sabry. The model specializes in understanding and translating ancient Egyptian Hieroglyphs into various modern human languages. In addition, it is highly multilingual and masters <b>150 Languages + Hieroglyphic</b>. To ensure accessibility and efficient deployment across different hardware setups, the model is also available in multiple GGUF variants.
</p>

## Model Overview
**Horus Hiero 4B Mini** is a 4-billion parameter language model built on top of the robust [Qwen/Qwen3.5-4B](https://huggingface.co/Qwen/Qwen3.5-4B) architecture (or its GGUF equivalent). It is designed to handle extremely long contexts and leverages a sophisticated hybrid attention mechanism to deliver high performance and efficiency.

### Leading Model in the Middle East

<img src="media/middle%20east%20top%20models.png" alt="Middle East Top Models" width="100%"/>

<br>

### 150 Languages + Hieroglyphic

<img src="media/languages.png" alt="Languages" width="100%"/>

<br>

### Example: Hieroglyph Translation

<img src="media/mode%20lchat.png" alt="Translation Example" width="100%"/>

## Model Variants
The following quantized versions are available for different deployment scenarios:

| Variant | Format | Size | Best For |
| :--- | :--- | :--- | :--- |
| **[Full 16-bit](https://huggingface.co/tokenaii/Horus-Hiero-Mini-4B)** | Safetensors | ~8.0 GB | Maximum quality, GPU inference |
| **[GGUF Repo](https://huggingface.co/tokenaii/Horus-Hiero-Mini-4B-GGUF)** | GGUF | Various | All quantized variants (Q2_K through Q8_0) |
| **[Q8_0](https://huggingface.co/tokenaii/Horus-Hiero-Mini-4B-GGUF/blob/main/Horus-Hiero-Mini-4B-Q8_0.gguf)** | GGUF | ~4.4 GB | Minimal quality loss |
| **[Q6_K](https://huggingface.co/tokenaii/Horus-Hiero-Mini-4B-GGUF/blob/main/Horus-Hiero-Mini-4B-Q6_K.gguf)** | GGUF | ~3.4 GB | Near-full quality |
| **[Q4_K_M](https://huggingface.co/tokenaii/Horus-Hiero-Mini-4B-GGUF/blob/main/Horus-Hiero-Mini-4B-Q4_K_M.gguf)** | GGUF | ~2.6 GB | Balanced quality/size, CPU+GPU |
| **[Q2_K](https://huggingface.co/tokenaii/Horus-Hiero-Mini-4B-GGUF/blob/main/Horus-Hiero-Mini-4B-Q2_K.gguf)** | GGUF | ~1.8 GB | Maximum compression, lower RAM usage |

## Technical Specifications

| Feature | Specification |
| :--- | :--- |
| **Number of Parameters** | 4 Billion |
| **Input Modalities** | Text, Image, Video |
| **Context Length** | 524,288 tokens (512K) natively |
| **Training** | Multi-Token Prediction (MTP), strong-to-weak distillation |
| **Thinking Mode** | Enabled (toggleable via `enable_thinking` parameter) |
## Benchmark Results

| Benchmark | Score |
| :--- | :---: |
| **MMLU-Pro** | **74.2%** |
| **GPQA Diamond** | **72.3%** |
| **HumanEval** | **68.2%** |
| **LiveCodeBench** | **51.4%** |
| **MATH** | **45.8%** |

## Quick Start
**Using NeuralNode (Recommended)**  
The easiest way to use Horus models is with the NeuralNode framework:

```python
import neuralnode as nn

# For GGUF models (recommended):
MODEL_ID = "tokenaii/Horus-Hiero-Mini-4B-GGUF/Horus-Hiero-Mini-4B-Q6_K.gguf"
DEVICE = "cpu"  # Change to "cuda" for GPU acceleration

# Download and load
model = nn.HorusModel(MODEL_ID, device=DEVICE).load()

# Use immediately
response = model.chat([
    {"role": "user", "content": "Translate this hieroglyph: 𓂋𓏏𓈖𓀀"}
])
print(response.content)
```

<br><p align="center"><a href="https://tokenai.llc/projects/horus-hiero"><b>Create your own code to use Horus Hiero →</b></a></p>

## About TokenAI & The Horus Family
**TokenAI** is a non-profit AI startup founded in 2025 by Assem Sabry, located in Alexandria, Egypt. 

The **Horus** family is our line of advanced language models. The series began with **Horus 1.0 4B**, which achieved remarkable success as the very first language model to be fully trained from scratch in Egypt. Building on that foundation, Horus Hiero brings specialized capabilities in ancient languages while retaining powerful modern multilingual performance.

### Contact & Community
- **Website:** [tokenai.llc](https://tokenai.llc/)
- **Hugging Face:** [tokenaii](https://huggingface.co/tokenaii)
- **GitHub:** [tokenaii](https://github.com/tokenaii)
- **LinkedIn:** [TokenAI](https://www.linkedin.com/company/tokenaii)
- **Location:** Alexandria, Egypt
## Citation
If you use Horus Hiero 4B Mini in your research or applications, please cite it as:
```bibtex
@misc{tokenai_horus_hiero_4b_mini,
    title={Horus Hiero 4B Mini: Advanced Multilingual and Hieroglyphs Language Model},
    author={Assem Sabry and TokenAI},
    year={2026},
    url={https://huggingface.co/tokenaii/Horus-Hiero-Mini-4B}
}
```