OpenBMB

VoxCPM2: Tokenizer-Free TTS for Multilingual Speech Generation, Creative Voice Design, and True-to-Life Cloning

ChatDev 2.0: Dev All through LLM-powered Multi-Agent Collaboration

See, hear, and interact on your local devices.

MiniCPM5: SOTA on-device LLMs, small yet powerful.

An Autonomous LLM Agent for Complex Task Solving

[ICLR'24 spotlight] An open platform for training, serving, and evaluating large language model for tool learning.

A Low-Code MCP Framework for Building Complex and Innovative RAG Pipelines

🤖 AgentVerse 🪐 is designed to facilitate the deployment of multiple LLM-based agents in various applications, which primarily provides two frameworks: task-solving and simulation

Task-oriented AI Agent productivity platform

Tool Learning for Big Models, Open-Source Solutions of ChatGPT-Plugins

百亿参数的中英文双语基座大模型

Enterprise Digital Employee Platform

AgentCPM-GUI: An on-device GUI agent for operating Android apps, enhancing reasoning ability with reinforcement fine-tuning for efficient task execution.

[ICLR'24 spotlight] Chinese and English Multimodal Large Model Series (Chat and Paint) | 基于CPM基础模型的中英双语多模态大模型系列

An LLM-powered repository agent designed to assist developers and teams in generating documentation and understanding repositories quickly.

Parsing-free RAG supported by VLMs

Fast, accurate OCR with correctness assessment.

An LLM-based Agent for the New Automation Paradigm - Agentic Process Automation

An open-source framework for collaborative AI agents, enabling diverse, distributed agents to team up and tackle complex tasks through internet-like connectivity.

An End-to-End Infrastructure for Training and Evaluating Various LLM Agents

Efficient Training (including pre-training and fine-tuning) for Big Models

Efficient Inference for Big Models

DeepThinkVLA: Enhancing Reasoning Capability of Vision-Language-Action Models

MiniCPM-V apps — fully offline multimodal chat on iOS / Android / HarmonyOS

Live Training for Open-source Big Models

a local-first desktop pet powered by MiniCPM5

Official PyTorch+CUDA Full-functional Web Demo for MiniCPM-o 4.5

Codes for the paper "∞Bench: Extending Long Context Evaluation Beyond 100K Tokens": https://arxiv.org/abs/2402.13718

A large-scale, fine-grained, diverse preference dataset (and models).

A Smarter and Faster On-Device AI Brain for Robots

A List of Big Models

Your faithful, impartial partner for audio evaluation — know yourself, know your rivals. 真实评测，知己知彼。A unified benchmark framework for ASR/TTS/Audio Codec/audio LLM evaluation

A collection of phenomenons observed during the scaling of big foundation models, which may be developed into consensus, principles, or laws in the future

Efficient, Low-Resource, Distributed transformer implementation based on BMTrain

[ACL 2024 Demo] Official GitHub repo for UltraEval: An open source framework for evaluating foundation models.

A General, Accurate, Long-Horizon, and Efficient Mobile Agent driven by Multimodal Foundation Models

CPM.cu is a lightweight, high-performance CUDA implementation for LLMs, optimized for end-device inference and featuring cutting-edge techniques in sparse architecture, speculative sampling and...

Extrapolating RLVR to General Domains without Verifiers

[ACL 2024]Official GitHub repo for OlympiadBench: A Challenging Benchmark for Promoting AGI with Olympiad-Level Bilingual Multimodal Scientific Problems.

An open platform for enhancing the capability of LLMs in workflow orchestration.

Model Compression for Big Models

SimpleNav: A simple, unified, reproducible, and extensible framework for navigation VLA research.

Native-video memory for vision-language-action models, using timestamped visual history and exact streaming inference for long-horizon robot manipulation.

A Toolkit for Running On-device Large Language Models (LLMs) in APP

Repo for paper "Tell Me More! Towards Implicit User Intention Understanding of Language Model Driven Agents"

[NeurIPS 2025] ParamMute (a.k.a. PIP-KAG): suppressing knowledge-critical FFNs for faithful retrieval-augmented generation (RAG knowledge conflicts)

ArcLight: A Lightweight LLM Inference Framework

Source code for ACL 2023 paper Decoder Tuning: Efficient Language Understanding as Decoding

ClawXMemory: A Multi-Level Memory Plugin for OpenClaw with Long-Term Context