transcribe_audio
Transcribe audio files to text with support for 32+ languages. Specify the language and provide a file path or URL for accurate transcription.
Instructions
Transcribe an audio file to text using Smallest AI's Pulse STT. Supports 32+ languages. IMPORTANT: Always ask the user what language the audio is in before calling this tool. Pass a file path on the user's machine (e.g. ~/Desktop/recording.wav) or a publicly accessible URL. Note: files uploaded to the chat sandbox are NOT accessible — ask the user for the actual file path on their machine or a URL instead.
Input Schema
TableJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
| file_path | No | Path to audio file on the user's machine (e.g. ~/Desktop/recording.wav, /Users/name/audio.mp3). NOT sandbox paths. Either file_path or audio_url is required. | |
| audio_url | No | Publicly accessible URL of an audio file. Either file_path or audio_url is required. | |
| language | Yes | Language of the audio. REQUIRED — ask the user. Use ISO 639-1 codes: en, hi, es, de, fr, it, pt, ta, mr, gu, bn, kn, ml, te, pa, or, ru, uk, pl, nl, sv, etc. Use 'multi' only if the user explicitly says they don't know the language. | |
| word_timestamps | No | Include word-level timestamps with confidence scores | |
| diarize | No | Enable speaker diarization (identify different speakers) | |
| emotion_detection | No | Detect emotions in speech | |
| redact_pii | No | Redact personally identifiable information from transcription |