Connecting AI,Humans, and the Physical World
Powered by multimodal models, bringing intelligence beyond the chat window and into the real world.
From models to products for everyday use
Talk to AI in real time. Choose a voice to start a call.
Try MOSS Live
Dialogue · AI assistant
A clear, friendly young male voice with a natural, conversational tone, suited to voice assistants, in-car interactions, and emotional companionship.
Narration · Natural female voice
A warm, natural, expressive female narrator who brings stories in children's picture books and audiobooks to life.
Character · Chinese male voice
A cool, reserved young male voice with a grounded, restrained delivery, suited to dramatic scenes in short-form and comic-based dramas.
Press Fn to speak in any app. Your text appears at the cursor.
Download MosMoscurl -X POST "https://api.mosi.cn/v1/audio/speech" \
-H "Authorization: Bearer $MOSS_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "moss-tts-1.5-flash",
"input": "Welcome to MOSS API. [pause 0.8s] Let us start generating speech.",
"voice_id": "94aa4989-c7e9-5007-ae42-ab401823e6c9",
"response_format": "mp3",
"delivery_method": "url"
}'Bring MOSS models to your product in a few lines of code
Get an API keyMOSS models for human-like interaction
MOSS-Transcribe-Diarize
Get transcripts, speaker identities, and timestamps together.
MOSS-TTS
Zero-shot voice cloning, reliable long-form speech, and control over duration and pronunciation.
MOSS-VL-Realtime
Instant answers, immediate corrections, and intentional silence in continuous video streams.
MOVA
Natively generate high-fidelity video with synchronized audio.
MOSS models for human-like interaction
Current model: MOSS-Transcribe-Diarize
