supertonic
Lightning-Fast, On-Device, Multilingual TTS — running natively via ONNX.
Discover how the Supertonic speed parameter (0.7-2.0) impacts audio latency and quality. Learn how to balance natural speech with computational speed for optimal results.
How to Migrate from Supertonic v2 to v3 While Maintaining ONNX Interface CompatibilityMigrate Supertonic v2 to v3 seamlessly. Maintain ONNX interface compatibility by swapping model and config files. Learn how to upgrade without inference code changes.
Supertonic vs VoxCPM2 WER/CER: How the 99M Parameter Model Matches 2B Parameter AccuracyDiscover how Supertonic's 99M parameter TTS model rivals VoxCPM2's 2B parameter accuracy in WER/CER. Get comparable or better results with fewer resources.
How Supertonic Preprocesses Text: Unicode Normalization and Emoji RemovalDiscover how Supertonic preprocesses text with its UnicodeProcessor. Learn about Unicode normalization, emoji removal, and a 10-step pipeline for clean, tokenized input.
How to Optimize Supertonic for Low‑Latency Real‑Time Synthesis on CPUOptimize Supertonic for low-latency CPU synthesis. Configure ONNX Runtime, reduce steps to 6-8, boost speed to 1.2-1.5, and use singleton sessions for minimal overhead.
How Supertonic’s ONNX Inference Pipeline Works: Duration Predictor, Text Encoder, Vector Estimator, and Vocoder ExplainedExplore Supertonic's ONNX inference pipeline. Understand how duration predictor, text encoder, vector estimator, and vocoder work together for advanced text to speech generation.
Required ONNX Model Files and How load_onnx_all() Loads Them in SupertonicDiscover the four essential ONNX model files required by Supertonic and learn how load_onnx_all() efficiently loads them using onnxruntime for your AI audio projects.
How to Set Up the Supertonic Serve HTTP Server with Custom EndpointsLearn to set up the supertonic serve HTTP server with custom endpoints. Extend built-in TTS functionality with your own business logic using FastAPI and APIRouter.
Supertonic Text-to-Speech Output Audio Format: 44.1 kHz 16-bit WAV ConfigurationDiscover Supertonic's default output audio format: 44.1 kHz 16-bit WAV mono files. Configure your audio settings for high-fidelity text-to-speech. Optimize your sound projects with Supertonic.
Maximum Input Text Length and Text Chunking in Supertonic: Implementation GuideDiscover Supertonic's maximum input text length limits and how it intelligently chunks long text into 300-character segments by default using sentence boundaries and a consistent algorithm.
How Supertonic Handles Financial Expressions, Phone Numbers, and Technical UnitsDiscover how Supertonic's TTS pipeline automatically handles financial expressions, phone numbers, and technical units with its built-in text normalizer. Get natural spoken output without pre-processing.
How to Load and Switch Between Voice Style Presets (M1‑M5, F1‑F5) in SupertonicLearn to load and switch Supertonic voice style presets M1-M5 and F1-F5 using loadVoiceStyle or load_voice_style. Easily manage voice styles at runtime by updating file paths or batch indexes.
Have a question about this repo?
These articles cover the highlights, but your codebase questions are specific. Give your agent direct access to the source. Share this with your agent to get started:
curl -s "https://instagit.com/install.md" 
Maintain an open-source project? Get it listed too →