Community
MOSS team holds Reddit AMA on its speech, audio, image and video model family
The Shanghai-based MOSS team, tied to OpenMOSS at Fudan University, is answering questions on r/LocalLLaMA. MOSS-Image is a 5B DiT text-to-image model under Apache 2.0, with base and SFT weights available.
- LabHugging Face
- ModelMOSS-Image
- ModelMOSS-TTS
- ModelMOSS-TTSD
- ModelMOSS-TTS-Realtime
- ModelMOSS-TTS-Nano
- ModelMOSS-VoiceGenerator
- ModelMOSS-SoundEffect
- ModelMOSS-Transcribe-Diarize
- ModelMOSS-Audio
- ModelMOSS-Music
- ModelMOSS-VL
- ModelMOSS-VL-Realtime
- ModelMOVA
- Topicaudio
- Topicimage
- Topicvideo
- Topicopen-weights
- Topicresearch
Every outlet that ran it 1 article
r/LocalLLaMACommunityLead
We’re MOSS, the team behind MOSS-Image, MOSS-TTS, and more — Ask Us Anything!