Skip to content

v0.3.0 — Batch Transcription Support

This release adds batch transcription support for pre-recorded audio files and URLs, complementing the existing real-time streaming transcription.

Highlights

  • Batch transcription support for pre-recorded audio files and URLs

Added

  • Transcribe method for batch transcription of audio bytes (d64c4ea)
  • TranscribeFile method for batch transcription of local audio files (d64c4ea)
  • TranscribeURL method for batch transcription of remote audio URLs (d64c4ea)
  • Pre-recorded transcription conversion helpers for Deepgram responses (3fc08dd)

Tests

  • TranscribeFile and TranscribeURL conformance tests with Ferris Bueller test audio (35c593e)

Dependencies

  • Bump github.com/agentplexus/omnivoice from v0.3.0 to v0.4.1 (35c593e)

Documentation

  • README updated with batch transcription examples (fe09c20)

Quick Start

import (
    deepgramstt "github.com/agentplexus/omnivoice-deepgram/omnivoice/stt"
    "github.com/agentplexus/omnivoice/stt"
)

// Create STT provider
provider, err := deepgramstt.New(deepgramstt.WithAPIKey("your-api-key"))
if err != nil {
    log.Fatal(err)
}

// Transcribe a local file
result, err := provider.TranscribeFile(ctx, "audio.mp3", stt.TranscriptionConfig{
    Language: "en",
})
if err != nil {
    log.Fatal(err)
}
fmt.Println(result.Text)

// Transcribe from URL
result, err = provider.TranscribeURL(ctx, "https://example.com/audio.mp3", stt.TranscriptionConfig{
    Language: "en",
})

See the README for more examples.