LMNT Voice Provider

LMNT provides ultra-low-latency text-to-speech synthesis optimized for real-time voice applications. The Beluga AI provider uses the LMNT Speech API for synthesis, supporting configurable voice selection, output format, and speech speed.

Choose LMNT when you need the lowest possible TTS latency for real-time conversational agents. LMNT is purpose-built for speed, making it a strong choice for voice pipelines where response time is the primary concern. For broader voice variety or voice cloning, consider ElevenLabs or PlayHT.

Installation

import _ "github.com/lookatitude/beluga-ai/voice/tts/providers/lmnt"

The blank import registers the "lmnt" provider with the TTS registry.

Configuration

Field	Type	Default	Description
Voice	string	"lily"	LMNT voice identifier
Format	AudioFormat	—	Output audio format
Speed	float64	—	Speech rate multiplier (1.0 = normal)
Extra	—	—	See below

Extra Fields

Key	Type	Required	Description
api_key	string	Yes	LMNT API key
base_url	string	No	Override base URL

Basic Usage

package main

import (
    "context"
    "log"
    "os"

    "github.com/lookatitude/beluga-ai/voice/tts"
    _ "github.com/lookatitude/beluga-ai/voice/tts/providers/lmnt"
)

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

    engine, err := tts.New("lmnt", tts.Config{
        Voice: "lily",
        Extra: map[string]any{"api_key": os.Getenv("LMNT_API_KEY")},
    })
    if err != nil {
        log.Fatal(err)
    }

    audio, err := engine.Synthesize(ctx, "Hello, welcome to Beluga AI.")
    if err != nil {
        log.Fatal(err)
    }

    if err := os.WriteFile("output.wav", audio, 0644); err != nil {
        log.Fatal(err)
    }
}

Direct Construction

import "github.com/lookatitude/beluga-ai/voice/tts/providers/lmnt"

engine, err := lmnt.New(tts.Config{
    Voice: "lily",
    Extra: map[string]any{"api_key": os.Getenv("LMNT_API_KEY")},
})

Streaming

The streaming interface synthesizes each text chunk independently:

for chunk, err := range engine.SynthesizeStream(ctx, textStream) {
    if err != nil {
        log.Printf("error: %v", err)
        break
    }
    transport.Send(chunk)
}

FrameProcessor Integration

processor := tts.AsFrameProcessor(engine, 24000, tts.WithVoice("lily"))
pipeline := voice.Chain(sttProcessor, llmProcessor, processor)

Advanced Features

Per-Request Options

audio, err := engine.Synthesize(ctx, "Hello!",
    tts.WithVoice("daniel"),
    tts.WithSpeed(1.5),
)

Custom Endpoint

engine, err := tts.New("lmnt", tts.Config{
    Voice: "lily",
    Extra: map[string]any{
        "api_key":  os.Getenv("LMNT_API_KEY"),
        "base_url": "https://lmnt.internal.corp/v1",
    },
})