whereismyai
kokoro-onnx
kokoro-onnx is a text-to-speech library that runs Kokoro-TTS models using the ONNX runtime for local audio generation.
Kokoro-onnx provides an implementation of Kokoro-TTS that executes locally via the ONNX runtime. It utilizes model files like kokoro-v1.0.onnx and voices-v1.0.bin to convert input text into audio.wav files. The system supports multiple languages and offers a range of voices for speech synthesis. It is designed for lightweight deployment, with standard models weighing approximately 300MB and quantized versions around 80MB. Users interact with the library through Python, requiring the setup of a project folder and the inclusion of specific model dependencies.
kokoro-onnx at a glance
Key features
- Supports multiple languages for speech synthesis
- Near real-time performance on macOS M1 hardware
- Provides access to multiple voice options
- Lightweight model size of approximately 300MB
- Quantized model support for 80MB footprint
- Compatible with misaki g2p package for version 1.0
Use it for
- Generating audio files from text input
- Integrating lightweight local TTS into Python projects
- Creating speech output in offline environments
Alternatives to kokoro-onnx
- Kokoro-FastAPI — Generate hours of high quality speech with an OpenAI-compatible endpoint.
- sherpa-onnx — Runs complex audio AI tasks offline on local devices and embedded hardware.
- ESPnet — Provides reproducible recipes and hundreds of pretrained models for various speech tasks.
- audio.cpp — Runs modern audio models locally with high-performance, native C++ execution.
- Piper — Runs locally and offers fast neural text-to-speech synthesis.
- Silero Models — Fast, portable, and simple text-to-speech models for CPU and GPU usage.
- MOSS-TTS — Delivers zero-shot voice cloning and fine-grained pronunciation control.
- WeNet — Achieves state-of-the-art results on public speech datasets.
- Speaches — It acts as an Ollama equivalent for running local TTS and STT models.
- FireRedAudio — A unified model handling understanding and generation with decoupled continuous representations.
- AssemblyAI — Industry-leading speech models deliver high accuracy and low latency for voice applications.
- RealtimeTTS — Provides low-latency speech synthesis with a broad engine matrix and multiple backends.
All kokoro-onnx alternatives