Beschreibung
Bring your game to life with fully offline, real-time text-to-speech. No internet required, no API keys, no subscriptions — everything runs locally on the player’s machine.
Auvyra TTS is powered by Piper TTS and ONNX Runtime, delivering high-quality, natural-sounding speech directly inside Unreal Engine. Auvyra TTS gives you access to 153 high-quality voice models across 40 languages, featuring 129 distinct voices and over 2,600 speaker variants — all manageable directly from a built-in editor UI inside Project Settings.
The plugin exposes a clean, easy-to-use Blueprint API. Synthesize a line of dialogue with a single node. Stream long paragraphs sentence by sentence so the first word plays within seconds.
🔗 Resources:
📖 Documentation
Whether you're building NPC dialogue, accessibility narration, procedural announcements, interactive fiction, or any system that needs voice — Auvyra TTS handles it without a server, without a subscription, and without slowing your game thread. All synthesis runs on a dedicated background thread.
Features:
🔒 100% Offline — No internet required. No API keys. No subscriptions. No usage limits. Runs entirely on the player's machine.
🌍 153 Voice Models, 40 Languages — English, French, German, Spanish, Chinese, Arabic, Hindi, Japanese, Russian, Ukrainian, and 30 more. Download only what you need.
🧑🤝🧑 2,661 Speaker Variants — 21 multi-speaker models let you create dozens of distinct NPC voices from a single downloaded model.
⚡ Instant Speech Playback — Long text starts playing almost immediately by streaming audio in chunks, so users hear results within seconds.
🧩 Full Blueprint API — Every feature is exposed to Blueprint. No C++ required.
🧠 Zero Game Thread Impact — All synthesis runs on a dedicated background thread. Your frame rate is never affected.
🛠 Built-in Voice Manager — Browse, download, delete, and preview all 153 voices directly inside Project Settings. No external tools needed.
🔊 In-Editor Audio Preview — Hear any voice before using it in your game, directly from the Voice Manager UI.
🎮 3D Spatial Audio — Play speech at a world location, attached to a moving NPC, or as flat 2D audio. Full attenuation support.
🚀 Voice Preloading — Preload voice models so speech plays instantly the first time, with no delay.
🎭 Multi-Speaker Support — Access hundreds of distinct voices from a single model file. Select speaker by index.
⏹ Instant Cancel — Interrupt speech at any time. Start new dialogue immediately without waiting for current speech to finish.
📦 Ready-to-Ship Builds — Everything you need is included automatically. Works seamlessly in both Development and Shipping Builds — no extra setup required.
🎚 Quality Tiers — Choose from x_low, low, medium, and high quality per voice to balance audio quality against file size and synthesis speed.
⚙️ Auto Packaging Setup — Plugin automatically configures your project's packaging settings on first editor launch. No manual config needed.