Shypd.aiThe largest open index of AI toolsFun-ASR is an end-to-end speech recognition large model launched by Tongyi Lab. It offers low-latency real-time transcription across 31 languages and excels in recognizing professional terminology.
Trending
About
Fun-ASR is an end-to-end speech recognition large model developed by Tongyi Lab, trained on tens of millions of hours of real speech data. It provides powerful contextual understanding and industry adaptability, supporting low-latency real-time transcription across 31 languages. The model is particularly adept at recognizing professional terminology and industry-specific expressions in vertical domains like education and finance, effectively addressing challenges such as "hallucination" generation and language confusion. Fun-ASR also features robust performance in far-field and high-noise environments, supports various Chinese dialects and regional accents, and offers enhanced lyric recognition under music interference. It is a fundamental speech recognition toolkit that includes ASR, VAD, Punctuation Restoration, Language Models, Speaker Verification, Speaker Diarization, and multi-talker ASR.
Capabilities
Pricing & Plans
Open Source
Free
FAQs
Trending
