←
MOSS455 posts
MOSS avatar

MOSS

@MOSS_Global

Connecting AI, humans, and the physical world. TTS|ASR|VL |Audio|Video|Contextual Intelligence ⚡️Vibe with us: discord.gg/Wjku5AYyxx

◷ Joined January 2026▧ 117 Photos and videos
MOSS avatar
MOSS@MOSS_Global·Oct 19

MOSS-Image API is coming soon in 24hrs Join MOSS discord for FREE early access 👉 discord.gg/Wjku5AYyxx

0:04
MOSS avatar
MOSS@MOSS_Global·23h

MOSS-VL-Realtime responds as the video plays. In @1yoursashh 's demo, it reports when a bee leaves and returns. After the scene changes, it still answers a follow-up about the earlier footage. Try it on your own video feed via the MOSS API 👇

1:25
Nelly; avatar
⟳ MOSS retweeted
Nelly;@nrqa__·Nov 9

speaker diarization only gets tested when people start interrupting each other.. so i made it harder. 3 astronauts in identical suits with mirrored visors, so you can't see who's talking. the voice is the only clue i ran the 45s scene through MOSS-Transcribe-Diarize Pro by @MOSS_Global ➤ 28 lines, split into S01, S02, S03 by voice alone ➤ it caught the interruptions and the moments two people talk over each other ➤ start and end timestamps on every line ➤ works on the platform or through the API, with keyterms for custom names every line landed on the right speaker.. full demo in the video. try it here: platform.moss.cn/?utm_0=2mIB…

1:42
MOSS avatar
MOSS@MOSS_Global·Nov 13

Your next “what if?” deserves an audience. 👀 Introducing the MOSS Creator Partner Program 🌱 Create, share, and earn up to 2,000 credits for qualifying submissions. Make something only you would make. 👾 Join CPP: discord.gg/KeWrGJ89qm

Post attachment
Andy Dandy avatar
⟳ MOSS retweeted
Andy Dandy@the_andydandy·Nov 12

Recently tried @MOSS_Global's new product. In short, it's Typeless + Plaud software in one. Accuracy on par with Typeless. But the part that got me is the Plaud-style magic: the meeting ends and your notes are already to-dos. Dictation, daily to-dos, meeting notes, all in one voice butler. Use my code KYQ39M3T gets you 2+ months of full access. Big meeting days? This is for you.

1:12
Adina Yakup avatar
⟳ MOSS retweeted
Adina Yakup@AdinaYakup·Nov 12

Nex-N2.6 🔥 New multimodal agentic model from Shanghai Innovation Institute - Based on GLM-5.3-Flash - 90% less repetitive reasoning (16% >> 1.6%) - 100% tool list adherence - Apache 2.0

0:45
Jeethu Rao avatar
⟳ MOSS retweeted
Jeethu Rao@jeethu·Nov 12

MOSS-Image-SFT with @_akhaliq’s 4-step LoRA generating an image in 7.02s on an iPhone 18 Pro Max.

0:10

MOSS-Image-SFT in just 4 steps. ⚡ @_akhaliq 's community distillation has reached 16.5K+ downloads on Hugging Face. Thanks for building and sharing! Try it on your own prompts ↓

Post attachment
MOSS avatar
MOSS@MOSS_Global·Nov 12

Watch Nex-N2.6 turn a single keyboard photo into an animated 3D showcase in Blender. ⌨️ Try it for free via the MOSS API 👇

0:45
MOSS avatar
MOSS@MOSS_Global·Nov 12

Meet the team behind MOSS-Image. The @ComfyUI -Org/MOSS-Image repo on Hugging Face is over 6.3M downloads in the past month. See what the model they trained can do. 👇

1:47