OpenMOSS (SII) ORG

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

MOSS-VL is the core multimodal model series within the OpenMOSS ecosystem, dedicated to visual understanding.

MOVA: Towards Scalable and Synchronized Video–Audio Generation

MOSS-TTS-Nano is an open-source multilingual tiny speech generation model from MOSI.AI and the OpenMOSS team. With only 0.1B parameters, it is designed for realtime speech generation, can run directly on CPU without a GPU, and keeps the deployment stack simple enough for local demos, web serving, and lightweight product integration.

Performant framework for training, analyzing and visualizing Sparse Autoencoders (SAEs) and their frontier variants.

MOSS-Transcribe-Diarize 0.9B is an open-source SOTA end-to-end audio understanding model for long-form multi-speaker transcription, diarization, timestamps, and acoustic event awareness.

MOSS‑TTS Family is an open‑source speech and sound generation model family from MOSI.AI and the OpenMOSS team. It is designed for high‑fidelity, high‑expressiveness, and complex real‑world scenarios, covering stable long‑form speech, multi‑speaker dialogue, voice/character design, environmental sound effects, and real‑time streaming TTS.

MOSS-TTSD is a spoken dialogue generation model designed for expressive multi-speaker synthesis. It features long-context modeling, flexible speaker control, and multilingual support, while enabling zero-shot voice cloning from short audio references.

A curated, continuously updated reading list, paper blogs, and resources for World Action Models (WAMs) in embodied AI.

Code for "AnyGPT: Unified Multimodal LLM with Discrete Sequence Modeling"

MOSS-Audio is an open-source foundation model for unified audio understanding, enabling speech, sound, music, captioning, QA, and reasoning in real-world scenarios.

Official repo of VLABench, a large scale benchmark designed for fairly evaluating VLA, Embodied Agent, and VLMs.

Collaborative Training of Large Language Models in an Efficient Way

GPT-4o-level, real-time spoken dialogue system.

MOSS-Audio-Tokenizer is a Causal Transformer-based audio tokenizer built on the CAT architecture. Trained on 3M hours of diverse audio, it supports streaming and variable bitrates, delivering SOTA reconstruction and strong performance in generation and understanding—serving as a unified interface for next-generation native audio language models.

[arXiv 26] FRoM-W1: Towards General Humanoid Whole-Body Control with Language Instructions

No description.

A real-time video understanding foundation model with gated cross-attention. Offline & real-time inference.

Dataset and evaluation script for "Evaluating Hallucinations in Chinese Large Language Models"

MOSS-Speech is a true speech-to-speech large language model without text guidance.

Official code of "RoboOmni: Proactive Robot Manipulation in Omni-modal Context"

MOSS-Music is an open-source music understanding model for targeting musical captioning, lyrics ASR, structural analysis, chord / key / tempo reasoning, and long-form musical question answering.

[ICML'2024] Can AI Assistants Know What They Don't Know?

[ACL'2024 Findings] GAOKAO-MM: A Chinese Human-Level Benchmark for Multimodal Models Evaluation

a survey of long-context LLMs from four perspectives, architecture, infrastructure, training, and evaluation

[AAAI26] LongLLaDA: Unlocking Long Context Capabilities in Diffusion LLMs

No description.

Official implementation of BandPO: Bridging Trust Regions and Ratio Clipping via Probability-Aware Bounds for LLM Reinforcement Learning. BandPO replaces canonical clipping (PPO/GRPO) with dynamic bounds to resolve exploration bottlenecks and prevent entropy collapse.

ABC-Bench is a benchmark for Agentic Backend Coding. It evaluates whether code agents can explore real repositories, edit code, configure environments, deploy containerized services, and pass external end-to-end API tests (HTTP-based integration tests) across realistic backend stacks.

[ICLR26] Beyond Real: Imaginary Extension of Rotary Position Embeddings for Long-Context LLMs

No description.

Drop-in async file-based handoff protocol for two AI coding agents (Claude Code + Codex), installed as one shared .handoff/ in your project.

No description.

Embodied-Planner-R1: Unleashing Embodied Task Planning Ability in LLMs via Reinforcement Learning

No description.

Institutional OpenClaw Solution. Share One Claw with Others.

VehicleWorld is the first comprehensive multi-device environment for intelligent vehicle interaction that accurately models the complex, interconnected systems in modern cockpits.

UnifiedToolHub is a comprehensive project supporting LLM-based tool use, designed to unify various tool-use dataset formats and provide training preparation, annotation, and evaluation functionalities.

[ICLR2025] ReAttention, a training-free approach to break the maximum context length in length extrapolation

No description.

No description.

No description.

A text-to-speech (TTS), speech-to-text (STT) and speech-to-speech (STS) library built on Apple's MLX framework, providing efficient speech analysis on Apple Silicon.

No description.

No description.

No description.

No description.

No description.

No description.

No description.

Ultra-Innerthought is a bilingual (Chinese and English) open-domain R1/o1 style SFT dataset.

No description.

A library for mechanistic interpretability of GPT-style language models

No description.

wangye