Tencent Hunyuan releases new speech recognition model Hy ASR3.0preview with multiple upgrades and free access to some capabilities

Tencent Hunyuan launched its next-generation speech recognition model Hy ASR3.0preview. The preview version is available on Hugging Face and ModelScope, and Tencent Hunyuan is offering free API quotas up to 1000 calls per day and free online use. The model supports up to 10 hours of extra-long audio, features robust noise resistance, and supports recognition in Mandarin, English, Japanese, Korean, and Cantonese. It also enables reasoning and illusion reduction, and can integrate with pure language models, aiming to achieve a new generation of speech intelligence.