Release Notes: v0.5.0

Release Date: 2026-06-14

Highlights

Added Gemini Live API support for native voice-to-voice processing with ~200ms latency.

What's New

Gemini Live API

The new omnivoice/realtime package provides native voice-to-voice capabilities via the Gemini Live API:

import "github.com/plexusone/omni-google/omnivoice/realtime"

provider := realtime.NewRealtimeProvider(os.Getenv("GEMINI_API_KEY"),
    realtime.WithModel("gemini-2.0-flash-live"),
    realtime.WithVoice("Puck"),
    realtime.WithInstructions("You are a helpful assistant."),
)

audioIn := make(chan []byte, 100)
audioCh, transcriptCh, err := provider.ProcessAudioStream(ctx, audioIn, corereal.ProcessConfig{})

Features:

• Low latency - ~200ms voice-to-voice response time
• Native audio - Model handles audio directly, no intermediate STT/TTS
• WebSocket streaming - Bidirectional audio via Gemini Live API
• Function calling - Execute tools during conversation with OnFunctionCall callback
• Multiple voices - Puck, Charon, Kore, Fenrir, Aoede

Audio Format

Direction	Format	Sample Rate	Channels
Input	PCM16	16kHz	Mono
Output	PCM16	24kHz	Mono

Configuration Options

realtime.WithModel("gemini-2.0-flash-live")
realtime.WithVoice("Puck")
realtime.WithInstructions("System prompt here")
realtime.WithTools(...)   // Function calling
realtime.WithTemperature(0.8)

Gateway Integration

Implements realtime.Provider interface from omnivoice-core for seamless gateway integration:

import (
    "github.com/plexusone/omnivoice-core/gateway"
    googleRealtime "github.com/plexusone/omni-google/omnivoice/realtime"
)

provider := googleRealtime.NewRealtimeProvider(os.Getenv("GEMINI_API_KEY"),
    googleRealtime.WithVoice("Puck"),
)

// Use with RealtimeBridge
bridge := gateway.NewRealtimeBridgeForTwilioGemini(provider, corereal.ProcessConfig{
    Instructions: "You are a helpful assistant.",
})

Dependencies

Package	Change
github.com/gorilla/websocket	Added v1.5.3
github.com/plexusone/omnivoice-core	Added v0.12.1

Getting Started

go get github.com/plexusone/omni-google@v0.5.0

See the Gemini Live API Guide for detailed usage.

Full Changelog

See CHANGELOG.md for the complete list of changes.