TencentHy ASR
Hy-ASR-3.0 Preview
Hy-ASR-3.0 Preview combines speech recognition with Hy3 language understanding. It supports Mandarin, English and twenty Chinese dialects.
Preview
Service and inputs
The preview is available through Tencent’s real-time speech interface. Some features remain unsupported in the preview. These limits describe the documented WebSocket engine; other media-processing integrations have separate request contracts.Tencent real-time speech recognition API
Practical use
Practical guidance. Include code-switching, accents and domain-specific names in your review. Semantic correction can improve readability but can also replace an unusual spoken term. Preserve the recording when exact wording matters.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Status | Preview |
| Access | API |
| Model identifiers | hy-asr-3.0-preview |
| Documented WebSocket input | 16 kHz mono PCM; audio under five minutesTencent real-time speech recognition API |