qwentts.cpp for Meari — Windows x64, CPU

Build of qwentts.cpp (MIT) used by Meari to run Qwen3-TTS 1.7B VoiceDesign (GGUF, Apache-2.0) without PyTorch — "make a new voice from a description".

Source: ServeurpersoCom/qwentts.cpp@5916bf7c79dbb4e66fc059cdc507a79a23a28bb4 (ggml cc5df49a) plus threads.patch (3 lines: read the CPU thread count from QWENTTS_N_THREADS, so the host can cap CPU use).
CPU only, all x86-64 variants (GGML_CPU_ALL_VARIANTS, GGML_BACKEND_DL): qwen-tts.exe picks the best ggml-cpu-*.dll for the machine at run time.
Includes the Microsoft C++ and OpenMP runtime DLLs (Visual Studio 2022 redistributable files) so it runs on a clean PC.
Build script: engine/tools/qwentts/build.ps1 in the Meari repository.

License: MIT — qwentts.cpp (LICENSE) and ggml (LICENSE-ggml). The model weights are separate (Apache-2.0, Serveurperso/Qwen3-TTS-GGUF).

Downloads last month: -; Downloads are not tracked for this model. How to track

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support