QwenAudio/SenseVoice Alternatives

a widely used multilingual ASR model; choose for larger language coverage and community, trading away emotion/event tags.

Robust Speech Recognition via Large-Scale Weak Supervision

a newer Qwen ASR model with timestamps and music/song recognition; choose if you want these features, but no emotion/event.

Qwen3-ASR is an open-source series of ASR models developed by the Qwen team at Alibaba Cloud, supporting stable multilingual speech/music/song recognition, language detection and timestamp prediction.

a current LLM-based ASR model family from Qwen; choose for streaming/vLLM support and larger capacity, without emotion/event.

Open-source LLM-based ASR model family for Chinese, dialect, accent, and multilingual speech, with FunASR, vLLM, streaming, and llama.cpp runtimes.

a full speech toolkit with its own pretrained models; choose it if you need training/inference plus TTS/speaker modules.

Easy-to-use Speech Toolkit including Self-Supervised Learning model, SOTA/Streaming ASR with punctuation, Streaming TTS with text frontend, Speaker Verification System, End-to-End Speech Translation and Keyword Spotting. Won NAACL2022 Best Demo Award.

a production streaming ASR toolkit; choose if streaming and deployability are your priority, not emotion/event.

Production First and Production Ready End-to-End Speech Recognition Toolkit

an enterprise speech framework with ASR/TTS models; choose if you are already in the NeMo ecosystem.

A scalable generative AI framework built for researchers and developers working on Large Language Models, Multimodal, and Speech AI (Automatic Speech Recognition and Text-to-Speech)

a PyTorch speech toolkit for research/training; choose to build custom systems, not for a single pretrained model.

A PyTorch-based Speech Toolkit

an end-to-end speech processing toolkit with many recipes; choose for research/custom training.

End-to-End Speech Processing Toolkit

a tiny multi-task audio model (ASR/TTS/VC); choose for extremely compact deployments, yet a different output set.

[AutoArk] GPA (General Purpose Audio) can do ASR, TTS and voice conversion with one tiny model!

a modular ASR+pipeline (VAD/LID/Punc); choose if you need a complete system for Chinese/English, not a single model.

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.

an ultra-low-latency on-device STT model; choose for edge voice agents, but it covers far fewer languages.

Very low latency speech to text, intent recognition, and text to speech, for building voice agents and interfaces

legacy training/deployment toolkit for English STT; choose only if you need to train your own models and can accept no maintenance.

🐸STT - The deep learning toolkit for Speech-to-Text. Training and deploying STT models has never been so easy.

an industrial ASR model for Mandarin/dialects/English; choose for Chinese-heavy workloads, but it lacks Japanese/Korean and emotion/event.

Open-source industrial-grade ASR models supporting Mandarin, Chinese dialects and English, achieving a new SOTA on public Mandarin ASR benchmarks, while also offering outstanding singing lyrics recognition capability.

a compact 1.5B ASR model; choose for a lightweight LLM-based alternative, likely with narrower language support.

GLM-ASR-Nano: A robust, open-source speech recognition model with 1.5B parameters

a GUI tool for transcription/subtitling/clipping; not a model to integrate.

FunASR-powered video transcription, subtitle generation, and LLM-assisted clipping tool with a local Gradio UI.