Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
A SOTA Industrial-Grade All-in-One ASR system with ASR, VAD, LID, and Punc modules. FireRedASR2 supports Chinese (Mandarin, 20+ dialects/accents), English, code-switching, and both speech and singing ASR. FireRedVAD supports speech/singing/music in 100+ langs. FireRedLID supports 100+ langs and 20+ zh dialects. FireRedPunc supports zh and en.
| Date | Stars |
|---|---|
| 2026-07-24 | 614 |
| 2026-07-25 | 615 |
| 2026-07-28 | 615 |
| 2026-07-30 | 615 |
| 2026-08-06 | 615 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.00%/day
<div align="center"> <h1> FireRedASR2S <br> A SOTA Industrial-Grade All-in-One ASR System </h1> </div> [[Paper]](https://arxiv.org/pdf/2603.10420) [[Model🤗]](https://huggingface.co/collections/FireRedTeam/fireredasr2s) [[Model🤖]](https://www.modelscope.cn/collections/xukaituo/FireRedASR2S) [[Demo]](https://huggingface.co/spaces/FireRedTeam/FireRedASR2S) FireRedASR2S is a state-of-the-art (SOTA), industrial-grade, all-in-one ASR system with ASR, VAD, LID, and Punc modules. All modules achieve SOTA performance: - **FireRedASR2**: Automatic Speech Recognition (ASR) supporting speech and singing transcription for Chinese (Mandarin, 20+ dialects/accents), English, code-switching. 2.89% average CER on 4 public Mandarin benchmarks, 11.55% on 19 Chinese dialects and accents benchmarks, **outperforming Doubao-ASR, Qwen3-ASR-1.7B, Fun-ASR, and Fun-ASR-Nano-2512**. FireRedASR2-AED also supports word-level timestamps and confidence scores. - **FireRedVAD**: Voice Activity Detection (VAD) supporting speech/singing/music in 100+ languages. 97.57% F1, **outperforming Silero-VAD, TEN-VAD, FunASR-VAD and WebRTC-VAD**. Supports non-streaming/streaming VAD and Multi-label VAD (mVAD). - **FireRedLID**: Spoken Language Identification (LID) supporting 100+ languages and 20+ Chinese dialects/accents. 97.18% accuracy, **outperforming Whisper and SpeechBrain**. - **FireRedPunc**: Punctuation Prediction (Punc) for Chinese and English. 78.90% average F1, outperforming FunASR-Punc (62.77%). *`2S`: `2`nd-generation FireRedASR, now expanded to an all-in-one ASR `S`ystem* ## 🔥 News - [2026.03.12] 🔥 We release FireRedASR2S technical report. See [arXiv](https://arxiv.org/abs/2603.10420). - [2026.03.05] 🚀 [vLLM](https://github.com/vllm-project/vllm/pull/35727) supports FireRedASR2-LLM. See [vLLM Usage](https://github.com/FireRedTeam/FireRedASR2S?tab=readme-ov-file#vllm-usage) part. - [2026.02.25] 🔥 We release **FireRedASR2-LLM model weights**. [🤗](https://huggingface.co/FireRedTeam/FireRedASR2-LLM) [🤖](https://www.modelscope.cn/models/xukaituo/FireRedASR2-LLM/) - [2026.02.13] 🚀 Support TensorRT-LLM inference acceleration for FireRedASR2-AED (contributed by NVIDIA). Benchmark on AISHELL-1 test set shows **12.7x speedup** over PyTorch baseline (single H20). - [2026.02.12] 🔥 We release FireRedASR2S (FireRedASR2-AED, FireRedVAD, FireRedLID, and FireRedPunc) with **model weights and inference code**. Download links below. Technical report and finetuning code coming soon. ## Available Models and Languages |Model|Supported Languages & Dialects|Download| |:-------------:|:---------------------------------:|:----------:| |FireRedASR2-LLM| Chinese (Mandarin and 20+ dialects/accents<sup>*</sup>), English, Code-Switching | [🤗](https://huggingface.co/FireRedTeam/FireRedASR2-LLM) \| [🤖](https://www.modelscope.cn/models/xukaituo/FireRedASR2-LLM/)| |FireRedASR2-AED| Chinese (Mandarin and 20+ dialects/accents<sup>*</sup>), English, Code-Switching | [🤗](https://huggingface.co/FireRedTeam/FireRedASR2-AED) \| [🤖](https://www.modelscope.cn/models/xukaituo/FireRedASR2-AED/)| |FireRedVAD | 100+ languages, 20+ Chinese dialects/accents<sup>*</sup> | [🤗](https://huggingface.co/FireRedTeam/FireRedVAD) \| [🤖](https://www.modelscope.cn/models/xukaituo/FireRedVAD/)| |FireRedLID | 100+ languages, 20+ Chinese dialects/accents<sup>*</sup> | [🤗](https://huggingface.co/FireRedTeam/FireRedLID) \| [🤖](https://www.modelscope.cn/models/xukaituo/FireRedLID/)| |FireRedPunc| Chinese, English | [🤗](https://huggingface.co/FireRedTeam/FireRedPunc) \| [🤖](https://www.modelscope.cn/models/xukaituo/FireRedPunc/)| <sup>*</sup>Supported Chinese dialects/accents: Cantonese (Hong Kong & Guangdong), Sichuan, Shanghai, Wu, Minnan, Anhui, Fujian, Gansu, Guizhou, Hebei, Henan, Hubei, Hunan, Jiangxi, Liaoning, Ningxia, Shaanxi, Shanxi, Shandong, Tianjin, Yunnan, etc. ## Evaluation ### FireRedASR2 Metrics: Character Error Rate (CER%) for Chinese and Word Error Rate (WER%) for English. Lower is better.
Excerpt of 24,088 characters
Read on GitHub
Kaituo Xu 许开拓 · Northwestern Polytechnical University · China
22
Emmanuel Schmidbauer
4
Yuekai Zhang · @Nvidia · China
2
2
Yifan Yang · Shanghai Jiao Tong University
1
1
Would you bet a product on this? Bounded 0–100 and slow moving.
matched fp:802ba6e3280800cd, topic:speech-recognition, topic:asr, readme:speech recognition
matched fp:802ba6e3280800cd, topic:llm, readme:model weights, readme:technical report
matched fp:802ba6e3280800cd, topic:multimodal-llm