VOICEVOX/voicevox

Features

• Text-to-Speech - Converts written text into high-quality spoken audio using a variety of synthetic character voices.
• Singing Voice Synthesis - Provides a specialized singing voice synthesizer with a piano-roll interface for melodic vocal composition.
• Voice - Allows selection of different voice characters and styles to change the identity of synthesized audio.
• Prosody Control - Offers fine-grained control over pitch, speed, and intonation by adjusting parameters tied to specific phonemes.
• Phoneme-Based Pipelines - Uses a customizable dictionary-based system to translate written text into phonetic representations for accurate pronunciation.
• Speech Synthesis Engines - Provides a core speech synthesis engine capable of being integrated into external applications or services.
• Synthetic Voice Design - Provides tools for creating and fine-tuning unique synthetic vocal identities and personas.
• Speech Processing Libraries - Provides a core library that allows speech generation capabilities to be embedded directly into external applications.
• Multi-Track Audio Editors - Ships a multi-track editor for organizing speech and singing segments with integrated synthesis.
• Multi-Track Audio Sequencers - Provides a multi-track sequencing environment to organize audio segments, timing, and character assignments on a project timeline.
• Vocal Compositions - Enables musical vocal composition using a multi-track editor with pitch editing and MIDI/UST imports.
• Hardware-Accelerated Inference - Supports switching between CPU and GPU processing to accelerate the speed of neural network audio generation.
• Phonetic Pronunciation Overrides - Supports custom pronunciation dictionaries for explicit phoneme sequences to resolve ambiguity.
• Speech Parameter Configuration - Allows configuration of auditory qualities like pitch, rate, and volume on a per-segment basis.
• Expressive Prosody Controls - Offers fine-grained controls for manipulating emotion, rhythm, and intonation.
• Hybrid Voice Synthesis - Blends characteristics of multiple target speakers to create unique hybrid vocal identities.
• Synthesis Project Management - Saves and loads work-in-progress sessions including text, character assignments, and timing settings.
• Docker Container Deployments - Packages the synthesis runtime into Docker containers to ensure consistent environment dependencies across operating systems.
• Standalone Service Deployments - Allows the synthesis engine to be deployed as a standalone executable or Docker container for external use.
• Humming Generation - Provides the ability to generate melodic humming sounds for characters to sing without lyrics.
• Client-Server Architectures - Implements a client-server architecture that decouples the user interface from the synthesis engine via a REST API.
• Voice Parameter Presets - Allows saving and applying groups of voice settings to maintain consistency across different segments of text.
• Voice Character Presets - Saves and loads collections of voice parameters to maintain consistent character delivery across different projects.

Star history