MongoDB - Build AI That Scales
Mini toolsMOSS TTS Nano
By OpenMOSS
MOSS-TTS-Nano is a compact multilingual text-to-speech model focused on practical deployment rather than maximum scale. The repository says it uses a pure autoregressive Audio Tokenizer plus LLM pipeline, supports streaming inference, runs on a 4-core CPU, handles long text with automatic chunked voice cloning, and outputs 48 kHz stereo audio. It currently supports 20 languages including Chinese, English, German, Spanish, Japanese, Korean, Arabic, Portuguese, and Turkish, and is released under Apache 2.0.
New
Multimodal
Gen 3
Overview
MOSS-TTS-Nano is an open-source multilingual tiny speech generation model from MOSI.AI and OpenMOSS. With only 0.1B parameters, it is built for real-time TTS, can run directly on CPU without a GPU, and keeps deployment simple enough for local demos, web serving, and lightweight product integration.
Tools using MOSS TTS Nano
No tools found for this model yet.
Last updated: April 14, 2026
