speech.detect-language

Language detection for audio routing uses the audio_transcribe tool to identify the primary spoken language. It returns the detected language, provider, duration, and cost, allowing you to select the correct transcription path without generating a full transcript.

The best way to route audio to the correct speech-to-text provider is by detecting the dominant spoken language first. This skill supports automatic provider selection based on the detected language to optimize downstream processing paths and costs.

Yes, you can identify the spoken language without generating a full transcript. The language detection process specifically avoids returning the full transcript unless explicitly requested, outputting only the detected language, provider, duration, cost, and warnings.

Yes, audio language detection supports automatic provider selection. It identifies the dominant spoken language and matches it with an appropriate transcription provider, returning the provider details, duration, and estimated cost for your routing decisions.

When detecting the language in an audio file, you receive the detected dominant language, selected provider, audio duration, cost estimation, and processing warnings. This routing-focused output provides the necessary metadata to configure downstream speech-to-text workflows.