---
title: Confucius4-R2T2-GGUF
canonical_url: "https://www.modelscope.cn/models/netease-youdao/Confucius4-R2T2-GGUF"
md_url: "https://www.modelscope.cn/models/netease-youdao/Confucius4-R2T2-GGUF.md"
repository: netease-youdao/Confucius4-R2T2-GGUF
last_updated: 2026-09-20
license: other
pipeline_tag: auto-speech-recognition
tasks:
  - auto-speech-recognition
base_model:
  - netease-youdao/Confucius4-R2T2
base_model_relation: quantized
library_name:
  - gguf
language:
  - zh
downloads: 633
stars: 3
tags:
  - GGUF
  - confucius4
  - r2t2
  - asr
  - streaming
  - real-time
  - low-latency
  - speech-recognition
  - multilingual
  - gguf
---

# Confucius4-R2T2-GGUF

> Confucius4-R2T2-GGUF - netease-youdao 在 ModelScope 开源的模型。Confucius4-R2T2: A Low Latency and High Accuracy Real-Time Speech Recognition Model Real Real-Time Transcription

netease-youdao/Confucius4-R2T2-GGUF 是 ModelScope 魔搭社区上的auto-speech-recognition模型，采用 other 许可，基于 netease-youdao/Confucius4-R2T2 构建。

- **Repository**: netease-youdao/Confucius4-R2T2-GGUF
- **License**: other
- **Tasks**: auto-speech-recognition
- **Base model**: netease-youdao/Confucius4-R2T2
- **Tags**: GGUF, confucius4, r2t2, asr, streaming, real-time, low-latency, speech-recognition, multilingual, gguf
- **Downloads**: 633
- **Stars**: 3
- **Last updated**: 2026-09-20

Source: https://www.modelscope.cn/models/netease-youdao/Confucius4-R2T2-GGUF

---

<div align="center">
    <img src="https://raw.githubusercontent.com/netease-youdao/Confucius4-R2T2/refs/heads/master/resources/R2T2_logo.png" alt="Confucius4-R2T2" width="30%">
    <h1>Confucius4-R2T2: A Low Latency and High Accuracy Real-Time Speech Recognition Model</h1>
        <p>
        <b>
            Real
            Real-Time
            Transcription
        </b>
    </p>
</div>

<div align="center">
    <a href="https://github.com/netease-youdao/Confucius4-R2T2"><img src="https://img.shields.io/badge/GitHub-Confucius4--R2T2-181717?logo=github" alt="GitHub repository"></a>
    &nbsp;&nbsp;&nbsp;&nbsp;
    <a href="https://github.com/netease-youdao/Confucius4-R2T2/blob/master/README.zh.md"><img src="https://img.shields.io/badge/README-中文版本-red" alt="Chinese README"></a>
    &nbsp;&nbsp;&nbsp;&nbsp;
    <a href="https://raw.githubusercontent.com/netease-youdao/Confucius4-R2T2/refs/heads/master/MODEL_LICENSE"><img src="https://img.shields.io/badge/model_license-NetEase-blue" alt="Model license: NetEase Model Use License Agreement"></a>
    &nbsp;&nbsp;&nbsp;&nbsp;
    <a href="https://github.com/netease-youdao/Confucius4-R2T2/blob/master/LICENSE"><img src="https://img.shields.io/badge/code_license-Apache%202.0-blue" alt="Code license: Apache 2.0"></a>
    &nbsp;&nbsp;&nbsp;&nbsp;
    <a href="https://r2t2.youdao.com/demo"><img src="https://img.shields.io/badge/Demo-在线体验-orange" alt="Online demo"></a>
    &nbsp;&nbsp;&nbsp;&nbsp;
    <a href="https://huggingface.co/netease-youdao/Confucius4-R2T2"><img src="https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-Confucius4R2T2-yellow" alt="Hugging Face model"></a>
    &nbsp;&nbsp;&nbsp;&nbsp;
    <a href="https://modelscope.cn/models/netease-youdao/Confucius4-R2T2"><img src="https://img.shields.io/badge/ModelScope-Confucius4R2T2-purple" alt="ModelScope model"></a>
    &nbsp;&nbsp;&nbsp;&nbsp;
    <a href="https://r2t2.ai/"><img src="https://img.shields.io/badge/Website-www.r2t2.ai-purple" alt="R2T2 website"></a>
    &nbsp;&nbsp;&nbsp;&nbsp;
</div>
<br>

## Confucius4-R2T2-GGUF

Official GGUF release from NetEase Youdao for [netease-youdao/Confucius4-R2T2](https://huggingface.co/netease-youdao/Confucius4-R2T2), a low-latency and high-accuracy true streaming Automatic Speech Recognition (ASR) model that features fine-grained and configurable decoding chunks from 80 ms to 2 s.

## Files

<table>
  <thead>
    <tr>
      <th>Filename</th>
      <th>Quantization</th>
      <th>Size</th>
      <th>Notes</th>
    </tr>
  </thead>
  <tbody>
    <tr>
      <td>Confucius4-R2T2-f16.gguf</td>
      <td>f16</td>
      <td>3.2GiB</td>
      <td>best quality</td>
    </tr>
    <tr>
      <td>Confucius4-R2T2-Q8_0.gguf</td>
      <td>Q8_0</td>
      <td>1.7GiB</td>
      <td>very good quality</td>
    </tr>
    <tr>
      <td>Confucius4-R2T2-Q4_K_M.gguf</td>
      <td>Q4_K_M</td>
      <td>1.0GiB</td>
      <td>fast, lower quality</td>
    </tr>
    <tr>
      <td>mmproj-Confucius4-R2T2-f16.gguf</td>
      <td>mmproj-f16</td>
      <td>0.6GiB</td>
      <td>multi-modal supplement</td>
    </tr>
    <tr>
      <td>mmproj-Confucius4-R2T2-Q8_0.gguf</td>
      <td>mmproj-Q8_0</td>
      <td>0.3GiB</td>
      <td>multi-modal supplement</td>
    </tr>
  </tbody>
</table>

The low-bit variants are quantized from the f16 GGUF.