---
title: cohere-transcribe-03-2026-demo
canonical_url: "https://www.modelscope.cn/studios/AI-ModelScope/cohere-transcribe-03-2026-demo"
md_url: "https://www.modelscope.cn/studios/AI-ModelScope/cohere-transcribe-03-2026-demo.md"
repository: AI-ModelScope/cohere-transcribe-03-2026-demo
chinese_name: "Cohere语音识别"
last_updated: 2026-06-16
sdk_type: gradio
sdk_version: 6.2.0
downloads: 0
stars: 14
---

# cohere-transcribe-03-2026-demo

> cohere-transcribe-03-2026-demo - AI-ModelScope 在 ModelScope 创建的在线 Demo。SOTA语音识别模型，支持14种语言，只需几秒钟就能将几分钟的音频转换成可用的转录文本

AI-ModelScope/cohere-transcribe-03-2026-demo 是 ModelScope 魔搭社区上的在线可交互 Demo（创空间），基于 gradio 6.2.0 构建，中文名为「Cohere语音识别」。

- **Repository**: AI-ModelScope/cohere-transcribe-03-2026-demo
- **SDK**: gradio
- **SDK version**: 6.2.0
- **Downloads**: 0
- **Stars**: 14
- **Last updated**: 2026-06-16

Source: https://www.modelscope.cn/studios/AI-ModelScope/cohere-transcribe-03-2026-demo

---

Evaluate multilingual speech recognition with a Gradio app backed by Cohere's compatibility transcription API.

## Local development

```bash
uv sync
COHERE_API_KEY=your-key uv run python app.py
```

Set `COHERE_ASR_DEBUG_RAW_RESPONSE=1` only when you want the raw API JSON panel to appear for debugging.

## Exporting Hugging Face dependencies

```bash
uv export --format requirements-txt --no-dev --no-hashes > requirements.txt
```

## Example audio

Bundled example clips live under `examples/` and are sourced from [`google/fleurs`](https://huggingface.co/datasets/google/fleurs) validation audio, one per supported language. The FLEURS dataset is licensed under CC-BY 4.0.