kokoro-onnx

Kokoro 82M voices as ONNX — real-time TTS on CPU with zero GPU.

Install Repository Compare

Fit.

Catalog snapshot

Platforms: Linux, Macos, Windows
Operating systems: Linux, macOS, Windows
Ram floor: 2 GB
Gpu: Not needed
Docker: No
Setup: Beginner
Deploy as: Local
Commercial use: Allowed

MIT wrapper; verify Kokoro weight licensing for your use case.

Project.

Fetched Sep 12, 2026¹

License: MIT
Stars: 2.7k¹
Forks: 302¹
Open issues: 73¹
Language: Python¹
Last commit: Sep 1, 2026¹
Tags: ttsonnxcpu

About.

kokoro-onnx packages the Kokoro 82M voice model as ONNX: real-time speech from a plain CPU with no server and no GPU. It is the lightest credible local TTS on a laptop or Pi. Model weights keep their own license terms.

82M model · ONNX runtime · CPU real-time · multiple voices

Alternatives.

Nearby.

All Speech AI

WWhisperOpenAI’s speech-to-text model you can run locally.Compare
PPiperFast local neural text-to-speech that runs on CPU.Compare
Wwhisper.cppC/C++ port of Whisper for fast local transcription.Compare
FFaster-WhisperWhisper reimplementation on CTranslate2 for faster local ASR.Compare
WwhisperXWhisper with word-level timestamps and speaker diarization for real subtitles.Compare