MoneyPrinterTurbo TTS Providers: The Complete List of Supported Engines

MoneyPrinterTurbo supports four TTS providers: Azure TTS v1, Azure TTS v2, SiliconFlow TTS, and Google Gemini TTS, selectable via the Audio Settings panel or programmatically through the app.services.voice module.

The open-source video generation framework harry0703/MoneyPrinterTurbo ships with a flexible text-to-speech architecture that accommodates multiple cloud and edge-based voice engines. All provider implementations reside in the app/services/voice.py module, exposing a unified interface for generating narration audio.

Supported TTS Providers in MoneyPrinterTurbo

MoneyPrinterTurbo integrates four distinct voice synthesis backends, each optimized for different latency, quality, and cost requirements:

• Azure TTS v1 (azure-tts-v1): A lightweight, synchronous wrapper around Microsoft's Edge TTS client (edge_tts.Communicate). Ideal for quick prototyping without Azure Speech SDK dependencies.
• Azure TTS v2 (azure-tts-v2): The full-featured Azure Speech SDK implementation (azure.cognitiveservices.speech) supporting word-boundary callbacks and enterprise-grade features.
• SiliconFlow TTS (siliconflow): HTTP-based integration with SiliconFlow's TTS API, supporting models like FunAudioLLM/CosyVoice2-0.5B.
• Google Gemini TTS (gemini-tts): Integration with Google's Gemini 2.5 Flash preview TTS model via the google.generativeai library.

The active provider is determined by the tts_server configuration value, which the web UI exposes as a dropdown selector.

How TTS Provider Selection Works

The selection mechanism operates at two layers: the user interface and the dispatch logic.

Web UI Configuration

In webui/Main.py (lines 48-53), the Audio Settings panel populates a selectbox with the four supported identifiers:

tts_servers = [
    ("azure-tts-v1", "Azure TTS V1"),
    ("azure-tts-v2", "Azure TTS V2"),
    ("siliconflow", "SiliconFlow TTS"),
    ("gemini-tts", "Google Gemini TTS"),
]

When a user selects a provider, the application stores the corresponding key in config.ui["tts_server"].

Backend Dispatch Logic

The voice.tts function in app/services/voice.py (lines 1127-1159) acts as a router. It inspects the configuration and delegates to the appropriate implementation:

# Simplified dispatch pattern from voice.py

if config.ui["tts_server"] == "azure-tts-v1":
    return azure_tts_v1(...)
elif config.ui["tts_server"] == "azure-tts-v2":
    return azure_tts_v2(...)
elif config.ui["tts_server"] == "siliconflow":
    return siliconflow_tts(...)
elif config.ui["tts_server"] == "gemini-tts":
    return gemini_tts(...)

This architecture allows code to remain provider-agnostic while supporting provider-specific optimizations.

Implementing TTS in Your Code

You can interact with MoneyPrinterTurbo's TTS system either through the high-level abstraction or by calling provider-specific functions directly.

Using the High-Level voice.tts Helper

The recommended approach uses the dispatch wrapper, which respects the user's UI selection:

from app.services import voice as vs

text = "Hello, world!"
voice_name = "en-US-JennyNeural"
voice_file = "/tmp/tts-output.mp3"
voice_rate = 1.0

sub_maker = vs.tts(
    text=text,
    voice_name=voice_name,
    voice_file=voice_file,
    voice_rate=voice_rate,
)

The tts function automatically routes to the configured provider based on config.ui["tts_server"].

Direct Provider Invocation

For workflows requiring specific provider features, import the individual functions from app/services/voice.py:

Azure TTS v1

sub = vs.azure_tts_v1(
    text="Welcome to Money Printer Turbo",
    voice_name="en-US-JennyNeural",
    voice_rate=1.0,
    voice_file="welcome-azure-v1.mp3",
)

Implementation reference: lines 1170-1199 of app/services/voice.py.

Azure TTS v2

sub = vs.azure_tts_v2(
    text="Welcome to Money Printer Turbo",
    voice_name="en-US-JennyMultilingualNeural",
    voice_file="welcome-azure-v2.mp3",
)

Implementation reference: lines 1343-1385 of app/services/voice.py.

SiliconFlow TTS

sub = vs.siliconflow_tts(
    text="Welcome to Money Printer Turbo",
    model="FunAudioLLM/CosyVoice2-0.5B",
    voice="FunAudioLLM/CosyVoice2-0.5B:alex",
    voice_file="welcome-siliconflow.mp3",
    voice_rate=1.0,
    voice_volume=1.0,
)

Implementation reference: lines 1205-1235 of app/services/voice.py.

Google Gemini TTS

sub = vs.gemini_tts(
    text="Welcome to Money Printer Turbo",
    voice="en-US-Standard-B",
    voice_rate=1.0,
    voice_volume=1.0,
    voice_file="welcome-gemini.mp3",
)

Implementation reference: lines 1438-1465 of app/services/voice.py.

Key Implementation Files

Understanding the file structure helps when extending or debugging TTS functionality:

• app/services/voice.py: Core implementation containing azure_tts_v1, azure_tts_v2, siliconflow_tts, gemini_tts, and the dispatch wrapper tts.
• webui/Main.py: UI layer defining the TTS provider dropdown (lines 48-53) that populates the tts_servers list.
• test/services/test_voice.py: Unit test coverage validating each provider's integration and audio generation pipeline.

Summary

• MoneyPrinterTurbo supports four TTS providers: Azure TTS v1, Azure TTS v2, SiliconFlow, and Google Gemini.
• Provider selection occurs via the Audio Settings dropdown in webui/Main.py, stored in config.ui["tts_server"].
• The voice.tts helper in app/services/voice.py (lines 1127-1159) automatically routes requests to the active provider.
• Direct invocation of provider-specific functions allows access to unique parameters like SiliconFlow's model selection or Gemini's voice identifiers.
• All implementations return a subtitle maker object (sub_maker) for synchronized caption generation.

Frequently Asked Questions

How do I switch between TTS providers in MoneyPrinterTurbo?

Navigate to the Audio Settings panel in the web UI and select your preferred engine from the TTS Server dropdown. Alternatively, programmatically set config.ui["tts_server"] to "azure-tts-v1", "azure-tts-v2", "siliconflow", or "gemini-tts" before calling voice.tts.

What is the difference between Azure TTS v1 and v2 in MoneyPrinterTurbo?

Azure TTS v1 uses the open-source edge_tts library for synchronous, edge-based synthesis without requiring Azure Speech SDK credentials. Azure TTS v2 leverages the official azure.cognitiveservices.speech SDK, supporting advanced features like word-boundary callbacks and enterprise speech resource management, but requires valid Azure subscription keys.

Can I use custom voice models with MoneyPrinterTurbo's TTS system?

Yes, but implementation varies by provider. SiliconFlow TTS explicitly supports custom model selection via the model parameter (e.g., FunAudioLLM/CosyVoice2-0.5B). For Azure providers, voice customization depends on Azure's available neural voices. The modular architecture in app/services/voice.py allows developers to add new provider functions following the existing pattern.

Where are the TTS provider configurations stored in the codebase?

The list of available providers is hardcoded in webui/Main.py lines 48-53 as the tts_servers tuple list. The active selection is stored in the runtime configuration dictionary under config.ui["tts_server"], which the dispatch logic in app/services/voice.py references to route requests to the appropriate implementation function.

{
  "mcpServers": {
    "instagit": {
      "command": "npx",
      "args": ["-y", "instagit@latest"]
    }
  }
}