Open source

Open Source Notices

Unitra AI is built with the help of these open source projects and their licenses.

Last updated: October 20, 2026

Libraries & Frameworks

Tauri MITApache-2.0
React MIT
Zustand MIT
Tailwind CSS MIT
i18next MIT
Lucide ISC
Tokio MIT
Serde MITApache-2.0
cpal Apache-2.0
Symphonia MPL-2.0

Runtime Components

sherpa-onnx Apache-2.0
ONNX Runtime MIT
DirectML Microsoft Software License Terms
llama.cpp / ggml MIT
LLVM OpenMP runtime Apache-2.0 WITH LLVM-exception

Model Attribution and Modifications

CosyVoice2 includes its language model, flow decoder, HiFT vocoder, CAM++ speaker encoder and speech tokenizer. Unitra converts the synthesis models to GGUF/ONNX as applicable and quantizes the language model without retraining.

Preset voice reference clips were generated offline with Qwen3-TTS 12Hz 0.6B CustomVoice. The generation model's weights are not distributed with Unitra.

The PP-OCRv6 small detection and recognition models are the publisher's ONNX exports, used without retraining or weight conversion. The Korean recognition pack uses PP-OCRv3. Lingua includes the 1.8.0 library and its bundled 1.3.0 language models.

The DNSMOS P.835 speech-quality model is created by Microsoft and published as part of the Deep Noise Suppression Challenge under CC-BY-4.0. Unitra uses sig_bak_ovr.onnx without modification, renamed on disk, to assess recorded speech while learning a voice profile. No endorsement by Microsoft is implied.

This page lists the open source components bundled with or downloaded by the Unitra AI desktop application. The license texts and attribution notices required for the AI models and runtime components above ship with the application and can be read in its settings. NVIDIA NeMo models are used under CC-BY-4.0, converted to ONNX and quantized without retraining; no endorsement by NVIDIA is implied. Trademarks are the property of their respective owners. For compliance inquiries: legal@unitra.ai.