MOSS-TTS v1.5
by OpenMOSS Expected Oct 13, 2026
Coming soonAudioopen-weightsttsthaivoice-cloningApache-2.0
thairouter/moss-tts-v1.5Overview
8B text-to-speech from Fudan's OpenMOSS lab covering 31 languages including Thai, with zero-shot voice cloning and 48 kHz stereo output.
Modality
Audio
text → speech
Licence
Apache-2.0
open weights
Pricing
TBA
per output, announced at launch
Weights
open
OpenMOSS/MOSS-TTS-v1.5
Expected Oct 13, 2026. Specs are provisional until the model is actually served; pricing is announced at launch.
Benchmarks
Reference scores published by the model's authors, not measured on ThaiRouter. Live numbers appear once the model is serving.
No published benchmark breakdown yet. Scores appear here as they are released.
API
Served through its own OpenAI-style endpoint, not chat completions.
Planned endpoint
https://api.thairouter.ai/v1/audio/speechSame API key, same credit balance, same no-training and no-storage guarantee as every text model. Request shape follows the OpenAI equivalent so existing SDK calls work; per-output pricing is announced at launch.
Not serving yet. Requests with this id return 404 until launch.