Skip to content

OmniVoice

Batteries-included voice pipeline framework for Go. Single import, all providers included.

Why OmniVoice?

  • 🎯 Single Import - One package for all STT, TTS, and telephony needs
  • 🗂️ Provider Registry - Get providers by name at runtime
  • 🔌 Multiple Providers - OpenAI, ElevenLabs, Deepgram, Twilio, Telnyx
  • ⚡ Streaming Support - Real-time transcription and synthesis
  • 📞 Voice Calls - Make and receive phone calls with CallSystem
  • 💬 SMS - Send text messages via SMSProvider
  • 🖥️ CLI Tool - Transcribe audio from the command line
  • 📄 Transcript Format - Canonical JSON format with timestamps and metadata
  • ⚡ Native Voice-to-Voice - OpenAI Realtime (~100ms) and Gemini Live (~200ms)

Quick Example

package main

import (
    "context"
    "log"
    "os"

    "github.com/plexusone/omnivoice"
    _ "github.com/plexusone/omnivoice/providers/all" // Register all providers
)

func main() {
    ctx := context.Background()

    // Get a TTS provider by name
    tts, err := omnivoice.GetTTSProvider("elevenlabs",
        omnivoice.WithAPIKey(os.Getenv("ELEVENLABS_API_KEY")))
    if err != nil {
        log.Fatal(err)
    }

    // Synthesize speech
    result, err := tts.Synthesize(ctx, "Hello, world!", omnivoice.SynthesisConfig{
        VoiceID: "pNInz6obpgDQGcFmaJgB",
    })
    if err != nil {
        log.Fatal(err)
    }

    // result.Audio contains the audio bytes
    log.Printf("Generated %d bytes of audio", len(result.Audio))
}

Installation

go get github.com/plexusone/omnivoice

Use Cases

Use Case Guide Providers
CLI Transcription CLI Guide All STT providers
Text-to-Speech TTS Guide OpenAI, ElevenLabs, Deepgram, Twilio
Speech-to-Text STT Guide OpenAI, ElevenLabs, Deepgram, Twilio
Voice Calls Voice Calls Twilio, Telnyx
SMS Messaging SMS Guide Twilio, Telnyx
Real-time Streaming Streaming All providers
Subtitles Subtitles All STT providers
Voice Agents Voice Agents Combined stack
Native Voice-to-Voice v0.9.0 Release OpenAI Realtime, Gemini Live

Architecture

┌─────────────────────────────────────────────────────────────────────────────┐
│                                  OmniVoice                                  │
│                            (batteries-included)                             │
├─────────────────────────────────────────────────────────────────────────────┤
│                                                                             │
│   import _ "github.com/plexusone/omnivoice/providers/all"                   │
│                                                                             │
│   ┌─────────────────────────────────────────────────────────────────────┐   │
│   │                           Provider Registry                         │   │
│   ├────────────┬────────────┬─────────────┬─────────────┬───────────────┤   │
│   │    TTS     │     STT    │  CallSystem │   Realtime  │  SMSProvider  │   │
│   ├────────────┼────────────┼─────────────┼─────────────┼───────────────┤   │
│   │ elevenlabs │ elevenlabs │   twilio    │ openai-     │    twilio     │   │
│   │ openai     │ openai     │   telnyx    │  realtime   │    telnyx     │   │
│   │ deepgram   │ deepgram   │             │ gemini-live │               │   │
│   │ twilio     │ twilio     │             │             │               │   │
│   └────────────┴────────────┴─────────────┴─────────────┴───────────────┘   │
│                                                                             │
└─────────────────────────────────────────────────────────────────────────────┘

Minimal Dependencies Alternative

For a minimal dependency footprint, use omnivoice-core with only the providers you need:

import (
    "github.com/plexusone/omnivoice-core/tts"
    elevenlabs "github.com/plexusone/elevenlabs-go/omnivoice/tts"
)

Next Steps