Audio Transcription Tool

This Tool enables Agents to transcribe speech from text using Audio Transcription Engines and Audio Transcription Drivers.

from griptape.drivers import OpenAiAudioTranscriptionDriver
from griptape.engines import AudioTranscriptionEngine
from griptape.structures import Agent
from griptape.tools.audio_transcription.tool import AudioTranscriptionTool

driver = OpenAiAudioTranscriptionDriver(model="whisper-1")

tool = AudioTranscriptionTool(
    off_prompt=False,
    engine=AudioTranscriptionEngine(
        audio_transcription_driver=driver,
    ),
)

Agent(tools=[tool]).run(
    "Transcribe the following audio file: /Users/andrew/code/griptape/tests/resources/sentences2.wav"
)