effect-uai basic-transcription

Finished audio in, text out. Use Transcriber.transcribe when the whole audio asset already exists.

Reach for this when the user says any of:

• "Transcribe this audio file / meeting / voice note"
• "I need word timestamps from audio"
• "Use OpenAI or Gemini for speech-to-text"

Pattern

import type { AudioSource } from "@effect-uai/core/Audio"
import * as Transcriber from "@effect-uai/core/Transcriber"