VNPT AI Ecosystem
Dedicated for
SmartVoice
API requested:
SmartVoice Engine Specifications
Technical specifications of the SmartVoice Engine when provided as a license.
Text to Speech
| Category | Specification |
|---|---|
| Integration Capability | API & SDK |
| Supported Protocols | gRPC, RestAPI |
| Minimum text length | 1 character/conversion |
| Maximum text length (synchronous) | 1,800 characters/conversion |
| Maximum text length (asynchronous) | 10,000 characters/conversion |
| Output | Link to audio file |
| Other Requirements | The converted voice has emotional intonation and expressiveness with male/female voices from 3 regional accents |
Speech to Text
| Category | Specification |
|---|---|
| Integration Capability | API & SDK |
| Supported Protocols | gRPC, RestAPI, websocket API gRPC |
| Voice Quality | Speaker's voice must be clearly audible, not too quiet (tested via mic at distance < 30cm) |
| Noise Environment | Environment without excessive background noise |
| File Format | Wav, Mp3 |
| Stream Format | PCM 16bit, Mono Channel |
| Synchronous audio file processing | File size < 10MB |
| Asynchronous audio file processing | File size < 250MB, maximum duration 2 hours |
| Format | PCM_16 (Int16 = 2bytes) |
| Sample rate | 16kHz |