ASR

To transcribe audio to text in a backend environment, use the provided scripts to process audio files or base64 input via the z-ai-web-dev-sdk, returning plain text. It supports single-file, batch, and directory transcription workflows.

Yes, you can use base64 input for speech-to-text processing by passing raw audio bytes to the backend SDK. This allows flexible transcription of audio data without needing direct file system access.

Yes, the z-ai-web-dev-sdk supports batch transcription for multiple audio files. You can process entire directories or batches of files with built-in error handling and metadata support.

The best way to automate speech-to-text transcription is integrating the SDK directly into your backend service or using command-line scripts for flexible workflows. This enables automated transcripts and voice-enabled applications.

You should not use this transcription approach for client-side applications because the SDK usage is restricted to server-side code. It is designed specifically for backend environments to ensure reliable processing.

No additional audio processing dependencies are needed beyond the z-ai-web-dev-sdk. It handles transcription directly from audio files or base64 input within your backend environment.