Repositories list

• MOSS-Transcribe-Diarize

  Public
  A 0.9B model for long-form transcription in 50+ languages with speaker diarization, timestamps, and acoustic event awareness

  Python

  •

  Apache License 2.0

  •106106 forks•1.8k1.8k stars•1616 issues•88 pull requests•Updated Sep 2, 2026Sep 2, 2026

• EasyWAM

  Public
  EasyWAM: A Unified and Efficient Framework for Training and Evaluating World Action Models.

  roboticscodebasefoundation-models+ 1

  roboticscodebasefoundation-modelsworld-action-model

  Python

  •

  Other

  •22 forks•4444 stars•11 issue•00 pull requests•Updated Sep 2, 2026Sep 2, 2026

• MOVA

  Public
  A foundation model that generates synchronized video and audio in a single model

  multimodaldiffusion-modelssglang+ 1

  multimodaldiffusion-modelssglangvideo-audio-generation

  Python

  •

  Apache License 2.0

  •9191 forks•1.1k1.1k stars•3030 issues•22 pull requests•Updated Sep 2, 2026Sep 2, 2026

• OmniVAE

  Public
  A unified audio-video variational autoencoder with cross-modal alignment for joint reconstruction and generation

  audiovideodit+ 5

  audiovideoditvideo-generationvaesaudio-generationmultimodal-generationaudio-video-generation

  Python

  •

  Apache License 2.0

  •33 forks•8484 stars•00 issues•00 pull requests•Updated Sep 2, 2026Sep 2, 2026

• Llamascopium

  Public
  A framework for training, analyzing, and visualizing sparse autoencoders and related interpretability methods

  sparse-autoencodersinterpretabilitysparse-dictionary+ 1

  sparse-autoencodersinterpretabilitysparse-dictionarymechanistic-interpretability

  Python

  •2929 forks•227227 stars•88 issues•00 pull requests•Updated Sep 2, 2026Sep 2, 2026

• .github

  Public
  OpenMOSS organization profile and shared GitHub community files

  00 forks•00 stars•00 issues•00 pull requests•Updated Sep 1, 2026Sep 1, 2026

• MOSS-VL

  Public
  An open-weight 11B model series for long-form and real-time video understanding

  visionstreaming-videovideo-understanding+ 10

  visionstreaming-videovideo-understandingvlmvision-languagellmsvision-language-modellong-video-understandingreal-time-aivision-language-models+ 3

  Python

  •

  Apache License 2.0

  •1818 forks•510510 stars•55 issues•00 pull requests•Updated Sep 1, 2026Sep 1, 2026

• Awesome-WAM

  Public
  A curated collection of papers, explainers, and resources on World Action Models for embodied AI

  HTML

  •

  MIT License

  •3636 forks•1.4k1.4k stars•44 issues•106106 pull requests•Updated Sep 1, 2026Sep 1, 2026

• SummerQuest-2026

  Public
  Public assignments and submissions for the 2026 OpenMOSS summer training program

  Python

  •6060 forks•88 stars•00 issues•1313 pull requests•Updated Aug 31, 2026Aug 31, 2026

• MOSS-TTS

  Public
  An open-source model family for long-form speech, dialogue synthesis, voice design, sound effects, and real-time streaming TTS

  audiotext-to-speechmultimodal+ 3

  audiotext-to-speechmultimodalvoice-cloningllmaudio-tokenizer

  Python

  •

  Apache License 2.0

  •367367 forks•4.1k4.1k stars•1414 issues•33 pull requests•Updated Aug 30, 2026Aug 30, 2026

• MOSS-TTSD

  Public
  A multilingual model for long-form, multi-speaker dialogue synthesis with flexible speaker control and zero-shot voice cloning

  streamingfinetunetext-to-speeh+ 3

  streamingfinetunetext-to-speehlarge-language-modelssglangspeech-dialogue-generation

  Python

  •

  Apache License 2.0

  •136136 forks•1.4k1.4k stars•5353 issues•00 pull requests•Updated Aug 30, 2026Aug 30, 2026

• MOSS

  Public
  An open-source, tool-augmented conversational language model from Fudan University

  natural-language-processingdeep-learningtext-generation+ 3

  natural-language-processingdeep-learningtext-generationdialogue-systemslarge-language-modelschatgpt

  Python

  •

  Apache License 2.0

  •1.1k1.1k forks•12k12k stars•236236 issues•77 pull requests•Updated Aug 30, 2026Aug 30, 2026

• MOSS-Audio

  Public
  An open-source model for understanding speech, environmental sounds, and music through captioning, question answering, and reasoning

  Python

  •4343 forks•654654 stars•1818 issues•11 pull request•Updated Aug 30, 2026Aug 30, 2026

• MOSS-TTS-Nano

  Public
  A 100M-parameter multilingual TTS model for real-time CPU inference, voice cloning, and 48 kHz stereo generation

  multilingualrealtimetts+ 6

  multilingualrealtimettsenglishchinesestreaming-audiomulti-modalityvoice-cloneaudio-tokenizer

  Python

  •

  Apache License 2.0

  •547547 forks•4.3k4.3k stars•1717 issues•77 pull requests•Updated Aug 30, 2026Aug 30, 2026

• SWE-bench-Science

  Public
  A benchmark for testing whether coding agents can resolve engineering tasks in scientific software

  Python

  •

  MIT License

  •11 fork•7575 stars•11 issue•00 pull requests•Updated Aug 25, 2026Aug 25, 2026

• OpenMOSS.github.io

  Public
  Source code for the OpenMOSS website and research project pages

  HTML

  •00 forks•22 stars•00 issues•00 pull requests•Updated Aug 23, 2026Aug 23, 2026

• OpenETA

  Public
  An embodied task agent that connects perception, action, verification, and learning in a continuous physical-world loop

  agentvlawam+ 6

  agentvlawamembodied-artificial-intelligenceembodied-agentembodied-aiembodied-intelligenceembodied-agentsphysical-ai

  Python

  •

  Apache License 2.0

  •1414 forks•167167 stars•11 issue•00 pull requests•Updated Aug 16, 2026Aug 16, 2026

• ContextWeave

  Public
  A benchmark for evaluating coding-agent memory systems through long-horizon personal worklog tasks

  Python

  •

  Other

  •00 forks•11 star•00 issues•00 pull requests•Updated Aug 6, 2026Aug 6, 2026

• AgentHPOBench

  Public
  A benchmark for evaluating LLM agents as sequential hyperparameter optimizers across 30 executable machine learning tasks

  Python

  •

  Apache License 2.0

  •00 forks•66 stars•00 issues•00 pull requests•Updated Aug 6, 2026Aug 6, 2026

• OmniVAE.github.io

  Public
  Project website for OmniVAE

  HTML

  •00 forks•00 stars•00 issues•00 pull requests•Updated Jul 27, 2026Jul 27, 2026

• MOSS-Video-Preview

  Public
  An 11B model for real-time video understanding with gated cross-attention and streaming inference

  Python

  •

  Apache License 2.0

  •44 forks•169169 stars•11 issue•00 pull requests•Updated Jul 16, 2026Jul 16, 2026

• claude-codex-handoff

  Public
  An asynchronous file-based handoff protocol for Claude Code and Codex agents sharing one project workspace

  opencodehandoffcodex+ 4

  opencodehandoffcodexai-agentsagent-collaborationclaude-codeclaude-skill

  Python

  •

  MIT License

  •11 fork•3737 stars•00 issues•00 pull requests•Updated Jul 4, 2026Jul 4, 2026

• MOSS-Audio-Tokenizer-Eval

  Public
  An objective reconstruction evaluation toolkit for audio tokenizers, neural codecs, audio VAEs, and vocoders

  Python

  •

  Apache License 2.0

  •00 forks•1313 stars•00 issues•00 pull requests•Updated Jun 16, 2026Jun 16, 2026

• MOSS-Audio-Tokenizer

  Public
  A 1.6B causal Transformer audio tokenizer with streaming, variable bitrates, and semantic alignment across speech, sound, and music

  audiomusictokenizer+ 4

  audiomusictokenizerspeechttsunifiedspeech-representation

  Python

  •

  Apache License 2.0

  •1818 forks•254254 stars•55 issues•11 pull request•Updated Jun 16, 2026Jun 16, 2026

• FRoM-W1

  Public
  Language-guided whole-body control for general humanoid robots

  whole-body-controlg1humanoid-robots+ 6

  whole-body-controlg1humanoid-robotsh1motion-generationfoundation-modelstext-to-motionunitreefftai

  Python

  •

  Apache License 2.0

  •99 forks•188188 stars•33 issues•00 pull requests•Updated Jun 5, 2026Jun 5, 2026

• sglang

  Public
  An OpenMOSS fork of SGLang with serving support for the MOSS-TTS family and MOSS-Audio

  Python

  •

  Apache License 2.0

  •00 forks•44 stars•00 issues•00 pull requests•Updated May 12, 2026May 12, 2026

• MOSS-VL-Demo

  Public
  Interactive web demo for MOSS-VL video understanding models

  Vue

  •00 forks•66 stars•00 issues•00 pull requests•Updated May 11, 2026May 11, 2026

• MOSS-Music

  Public
  An open-source model for music captioning, lyrics transcription, structural analysis, and musical question answering

  Python

  •1010 forks•150150 stars•33 issues•11 pull request•Updated May 9, 2026May 9, 2026

• MOSS-TTS-Nano-Reader

  Public
  A local browser reading app powered by MOSS-TTS-Nano with in-browser ONNX inference

  JavaScript

  •99 forks•5959 stars•00 issues•00 pull requests•Updated May 7, 2026May 7, 2026

• mlx-audio

  Public
  A library for text-to-speech, speech-to-text, and speech-to-speech inference on Apple Silicon with MLX

  Python

  •

  MIT License

  •702702 forks•66 stars•00 issues•00 pull requests•Updated Apr 27, 2026Apr 27, 2026