
MARBLE
Code for MultiAgentBench : Evaluating the Collaboration and Competition of LLM agents https://www.arxiv.org/pdf/2503.01935
Stars: 61

MARBLE (Multi-Agent Coordination Backbone with LLM Engine) is a modular framework for developing, testing, and evaluating multi-agent systems leveraging Large Language Models. It provides a structured environment for agents to interact in simulated environments, utilizing cognitive abilities and communication mechanisms for collaborative or competitive tasks. The framework features modular design, multi-agent support, LLM integration, shared memory, flexible environments, metrics and evaluation, industrial coding standards, and Docker support.
README:
Multi-Agent CooRdination Backbone with LLM Engine
MultiAgentBench is a modular and extensible framework designed to facilitate the development, testing, and evaluation of multi-agent systems leveraging Large Language Models (LLMs). It provides a structured environment where agents can interact within various simulated environments, utilizing cognitive abilities and communication mechanisms to perform tasks collaboratively or competitively.
- Modular Design: Easily extend or replace components like agents, environments, and LLM integrations.
- Multi-Agent Support: Model complex interactions between multiple agents with hierarchical or cooperative execution modes.
- LLM Integration: Interface with various LLM providers (OpenAI, etc.) through a unified API.
- Shared Memory: Implement shared memory mechanisms for agent communication and collaboration.
- Flexible Environments: Support for different simulated environments like web-based tasks.
- Metrics and Evaluation: Built-in evaluation metrics to assess agent performance on tasks.
- Industrial Coding Standards: High-quality, well-documented code adhering to industry best practices.
- Docker Support: Containerized setup for consistent deployment and easy experimentation.
Use a virtual environment, e.g. with anaconda3:
conda create -n marble python=3.10
conda activate marble
curl -sSL https://install.python-poetry.org | python3
export PATH="$HOME/.local/bin:$PATH"
Environment variables such as OPENAI_API_KEY
and Together_API_KEY
related configs are required to run the code. The recommended way to set all the required variable is
- Copy the
.env.template
file into the project root with the name.env
.
cp .env.template .env
- Fill the required environment variables in the
.env
file.
To run examples provided in the examples
:
poetry install
cd scripts
cd werewolf
bash run_simulation.sh
git checkout -b feature/feature-name
and PR to main
branch.
Run poetry run pytest
to make sure all tests pass (this will ensure dynamic typing passed with beartype) and poetry run mypy --config-file pyproject.toml .
to check static typing. (You can also run pre-commit run --all-files
to run all checks)
Check the github action result to make sure all tests pass. If not, fix the errors and push again.
Please cite the following paper if you find Marble helpful!
@misc{zhu2025multiagentbenchevaluatingcollaborationcompetition,
title={MultiAgentBench: Evaluating the Collaboration and Competition of LLM agents},
author={Kunlun Zhu and Hongyi Du and Zhaochen Hong and Xiaocheng Yang and Shuyi Guo and Zhe Wang and Zhenhailong Wang and Cheng Qian and Xiangru Tang and Heng Ji and Jiaxuan You},
year={2025},
eprint={2503.01935},
archivePrefix={arXiv},
primaryClass={cs.MA},
url={https://arxiv.org/abs/2503.01935},
}
For Tasks:
Click tags to check more tools for each tasksFor Jobs:
Alternative AI tools for MARBLE
Similar Open Source Tools

MARBLE
MARBLE (Multi-Agent Coordination Backbone with LLM Engine) is a modular framework for developing, testing, and evaluating multi-agent systems leveraging Large Language Models. It provides a structured environment for agents to interact in simulated environments, utilizing cognitive abilities and communication mechanisms for collaborative or competitive tasks. The framework features modular design, multi-agent support, LLM integration, shared memory, flexible environments, metrics and evaluation, industrial coding standards, and Docker support.

Vodalus-Expert-LLM-Forge
Vodalus Expert LLM Forge is a tool designed for crafting datasets and efficiently fine-tuning models using free open-source tools. It includes components for data generation, LLM interaction, RAG engine integration, model training, fine-tuning, and quantization. The tool is suitable for users at all levels and is accompanied by comprehensive documentation. Users can generate synthetic data, interact with LLMs, train models, and optimize performance for local execution. The tool provides detailed guides and instructions for setup, usage, and customization.

eole
EOLE is an open language modeling toolkit based on PyTorch. It aims to provide a research-friendly approach with a comprehensive yet compact and modular codebase for experimenting with various types of language models. The toolkit includes features such as versatile training and inference, dynamic data transforms, comprehensive large language model support, advanced quantization, efficient finetuning, flexible inference, and tensor parallelism. EOLE is a work in progress with ongoing enhancements in configuration management, command line entry points, reproducible recipes, core API simplification, and plans for further simplification, refactoring, inference server development, additional recipes, documentation enhancement, test coverage improvement, logging enhancements, and broader model support.

eliza
Eliza is a versatile AI agent operating system designed to support various models and connectors, enabling users to create chatbots, autonomous agents, handle business processes, create video game NPCs, and engage in trading. It offers multi-agent and room support, document ingestion and interaction, retrievable memory and document store, and extensibility to create custom actions and clients. Eliza is easy to use and provides a comprehensive solution for AI agent development.

comfyui_LLM_Polymath
LLM Polymath Chat Node is an advanced Chat Node for ComfyUI that integrates large language models to build text-driven applications and automate data processes, enhancing prompt responses by incorporating real-time web search, linked content extraction, and custom agent instructions. It supports both OpenAI’s GPT-like models and alternative models served via a local Ollama API. The core functionalities include Comfy Node Finder and Smart Assistant, along with additional agents like Flux Prompter, Custom Instructors, Python debugger, and scripter. The tool offers features for prompt processing, web search integration, model & API integration, custom instructions, image handling, logging & debugging, output compression, and more.

Mira
Mira is an agentic AI library designed for automating company research by gathering information from various sources like company websites, LinkedIn profiles, and Google Search. It utilizes a multi-agent architecture to collect and merge data points into a structured profile with confidence scores and clear source attribution. The core library is framework-agnostic and can be integrated into applications, pipelines, or custom workflows. Mira offers features such as real-time progress events, confidence scoring, company criteria matching, and built-in services for data gathering. The tool is suitable for users looking to streamline company research processes and enhance data collection efficiency.

VeritasGraph
VeritasGraph is an enterprise-grade graph RAG framework designed for secure, on-premise AI applications. It leverages a knowledge graph to perform complex, multi-hop reasoning, providing transparent, auditable reasoning paths with full source attribution. The framework excels at answering complex questions that traditional vector search engines struggle with, ensuring trust and reliability in enterprise AI. VeritasGraph offers full control over data and AI models, verifiable attribution for every claim, advanced graph reasoning capabilities, and open-source deployment with sovereignty and customization.

TaskingAI
TaskingAI brings Firebase's simplicity to **AI-native app development**. The platform enables the creation of GPTs-like multi-tenant applications using a wide range of LLMs from various providers. It features distinct, modular functions such as Inference, Retrieval, Assistant, and Tool, seamlessly integrated to enhance the development process. TaskingAI’s cohesive design ensures an efficient, intelligent, and user-friendly experience in AI application development.

Director
Director is a framework to build video agents that can reason through complex video tasks like search, editing, compilation, generation, etc. It enables users to summarize videos, search for specific moments, create clips instantly, integrate GenAI projects and APIs, add overlays, generate thumbnails, and more. Built on VideoDB's 'video-as-data' infrastructure, Director is perfect for developers, creators, and teams looking to simplify media workflows and unlock new possibilities.

griptape
Griptape is a modular Python framework for building AI-powered applications that securely connect to your enterprise data and APIs. It offers developers the ability to maintain control and flexibility at every step. Griptape's core components include Structures (Agents, Pipelines, and Workflows), Tasks, Tools, Memory (Conversation Memory, Task Memory, and Meta Memory), Drivers (Prompt and Embedding Drivers, Vector Store Drivers, Image Generation Drivers, Image Query Drivers, SQL Drivers, Web Scraper Drivers, and Conversation Memory Drivers), Engines (Query Engines, Extraction Engines, Summary Engines, Image Generation Engines, and Image Query Engines), and additional components (Rulesets, Loaders, Artifacts, Chunkers, and Tokenizers). Griptape enables developers to create AI-powered applications with ease and efficiency.

ai_automation_suggester
An integration for Home Assistant that leverages AI models to understand your unique home environment and propose intelligent automations. By analyzing your entities, devices, areas, and existing automations, the AI Automation Suggester helps you discover new, context-aware use cases you might not have considered, ultimately streamlining your home management and improving efficiency, comfort, and convenience. The tool acts as a personal automation consultant, providing actionable YAML-based automations that can save energy, improve security, enhance comfort, and reduce manual intervention. It turns the complexity of a large Home Assistant environment into actionable insights and tangible benefits.

Alice
Alice is an open-source AI companion designed to live on your desktop, providing voice interaction, intelligent context awareness, and powerful tooling. More than a chatbot, Alice is emotionally engaging and deeply useful, assisting with daily tasks and creative work. Key features include voice interaction with natural-sounding responses, memory and context management, vision and visual output capabilities, computer use tools, function calling for web search and task scheduling, wake word support, dedicated Chrome extension, and flexible settings interface. Technologies used include Vue.js, Electron, OpenAI, Go, hnswlib-node, and more. Alice is customizable and offers a dedicated Chrome extension, wake word support, and various tools for computer use and productivity tasks.

Zentara-Code
Zentara Code is an AI coding assistant for VS Code that turns chat instructions into precise, auditable changes in the codebase. It is optimized for speed, safety, and correctness through parallel execution, LSP semantics, and integrated runtime debugging. It offers features like parallel subagents, integrated LSP tools, and runtime debugging for efficient code modification and analysis.

CortexON
CortexON is an open-source, multi-agent AI system designed to automate and simplify everyday tasks. It integrates specialized agents like Web Agent, File Agent, Coder Agent, Executor Agent, and API Agent to accomplish user-defined objectives. CortexON excels at executing complex workflows, research tasks, technical operations, and business process automations by dynamically coordinating the agents' unique capabilities. It offers advanced research automation, multi-agent orchestration, integration with third-party APIs, code generation and execution, efficient file and data management, and personalized task execution for travel planning, market analysis, educational content creation, and business intelligence.

omniscient
Omniscient is an advanced AI Platform offered as a SaaS, empowering projects with cutting-edge artificial intelligence capabilities. Seamlessly integrating with Next.js 14, React, Typescript, and APIs like OpenAI and Replicate, it provides solutions for code generation, conversation simulation, image creation, music composition, and video generation.

llmc
llmc is an off-the-shell tool designed for compressing LLM, leveraging state-of-the-art compression algorithms to enhance efficiency and reduce model size without compromising performance. It provides users with the ability to quantize LLMs, choose from various compression algorithms, export transformed models for further optimization, and directly infer compressed models with a shallow memory footprint. The tool supports a range of model types and quantization algorithms, with ongoing development to include pruning techniques. Users can design their configurations for quantization and evaluation, with documentation and examples planned for future updates. llmc is a valuable resource for researchers working on post-training quantization of large language models.
For similar tasks

OSWorld
OSWorld is a benchmarking tool designed to evaluate multimodal agents for open-ended tasks in real computer environments. It provides a platform for running experiments, setting up virtual machines, and interacting with the environment using Python scripts. Users can install the tool on their desktop or server, manage dependencies with Conda, and run benchmark tasks. The tool supports actions like executing commands, checking for specific results, and evaluating agent performance. OSWorld aims to facilitate research in AI by providing a standardized environment for testing and comparing different agent baselines.

LLM-Agents-Papers
A repository that lists papers related to Large Language Model (LLM) based agents. The repository covers various topics including survey, planning, feedback & reflection, memory mechanism, role playing, game playing, tool usage & human-agent interaction, benchmark & evaluation, environment & platform, agent framework, multi-agent system, and agent fine-tuning. It provides a comprehensive collection of research papers on LLM-based agents, exploring different aspects of AI agent architectures and applications.

council
Council is an open-source platform designed for the rapid development and deployment of customized generative AI applications using teams of agents. It extends the LLM tool ecosystem by providing advanced control flow and scalable oversight for AI agents. Users can create sophisticated agents with predictable behavior by leveraging Council's powerful approach to control flow using Controllers, Filters, Evaluators, and Budgets. The framework allows for automated routing between agents, comparing, evaluating, and selecting the best results for a task. Council aims to facilitate packaging and deploying agents at scale on multiple platforms while enabling enterprise-grade monitoring and quality control.

ComfyBench
ComfyBench is a comprehensive benchmark tool designed to evaluate agents' ability to design collaborative AI systems in ComfyUI. It provides tasks for agents to learn from documents and create workflows, which are then converted into code for better understanding by LLMs. The tool measures performance based on pass rate and resolve rate, reflecting the correctness of workflow execution and task realization. ComfyAgent, a component of ComfyBench, autonomously designs new workflows by learning from existing ones, interpreting them as collaborative AI systems to complete given tasks.

MARBLE
MARBLE (Multi-Agent Coordination Backbone with LLM Engine) is a modular framework for developing, testing, and evaluating multi-agent systems leveraging Large Language Models. It provides a structured environment for agents to interact in simulated environments, utilizing cognitive abilities and communication mechanisms for collaborative or competitive tasks. The framework features modular design, multi-agent support, LLM integration, shared memory, flexible environments, metrics and evaluation, industrial coding standards, and Docker support.

Awesome-Agent-Papers
This repository is a comprehensive collection of research papers on Large Language Model (LLM) agents, organized across key categories including agent construction, collaboration mechanisms, evolution, tools, security, benchmarks, and applications. The taxonomy provides a structured framework for understanding the field of LLM agents, bridging fragmented research threads by highlighting connections between agent design principles and emergent behaviors.

LLM-Agent-Evaluation-Survey
LLM-Agent-Evaluation-Survey is a tool designed to gather feedback and evaluate the performance of AI agents. It provides a user-friendly interface for users to rate and provide comments on the interactions with AI agents. The tool aims to collect valuable insights to improve the AI agents' capabilities and enhance user experience. With LLM-Agent-Evaluation-Survey, users can easily assess the effectiveness and efficiency of AI agents in various scenarios, leading to better decision-making and optimization of AI systems.
For similar jobs

weave
Weave is a toolkit for developing Generative AI applications, built by Weights & Biases. With Weave, you can log and debug language model inputs, outputs, and traces; build rigorous, apples-to-apples evaluations for language model use cases; and organize all the information generated across the LLM workflow, from experimentation to evaluations to production. Weave aims to bring rigor, best-practices, and composability to the inherently experimental process of developing Generative AI software, without introducing cognitive overhead.

LLMStack
LLMStack is a no-code platform for building generative AI agents, workflows, and chatbots. It allows users to connect their own data, internal tools, and GPT-powered models without any coding experience. LLMStack can be deployed to the cloud or on-premise and can be accessed via HTTP API or triggered from Slack or Discord.

VisionCraft
The VisionCraft API is a free API for using over 100 different AI models. From images to sound.

kaito
Kaito is an operator that automates the AI/ML inference model deployment in a Kubernetes cluster. It manages large model files using container images, avoids tuning deployment parameters to fit GPU hardware by providing preset configurations, auto-provisions GPU nodes based on model requirements, and hosts large model images in the public Microsoft Container Registry (MCR) if the license allows. Using Kaito, the workflow of onboarding large AI inference models in Kubernetes is largely simplified.

PyRIT
PyRIT is an open access automation framework designed to empower security professionals and ML engineers to red team foundation models and their applications. It automates AI Red Teaming tasks to allow operators to focus on more complicated and time-consuming tasks and can also identify security harms such as misuse (e.g., malware generation, jailbreaking), and privacy harms (e.g., identity theft). The goal is to allow researchers to have a baseline of how well their model and entire inference pipeline is doing against different harm categories and to be able to compare that baseline to future iterations of their model. This allows them to have empirical data on how well their model is doing today, and detect any degradation of performance based on future improvements.

tabby
Tabby is a self-hosted AI coding assistant, offering an open-source and on-premises alternative to GitHub Copilot. It boasts several key features: * Self-contained, with no need for a DBMS or cloud service. * OpenAPI interface, easy to integrate with existing infrastructure (e.g Cloud IDE). * Supports consumer-grade GPUs.

spear
SPEAR (Simulator for Photorealistic Embodied AI Research) is a powerful tool for training embodied agents. It features 300 unique virtual indoor environments with 2,566 unique rooms and 17,234 unique objects that can be manipulated individually. Each environment is designed by a professional artist and features detailed geometry, photorealistic materials, and a unique floor plan and object layout. SPEAR is implemented as Unreal Engine assets and provides an OpenAI Gym interface for interacting with the environments via Python.

Magick
Magick is a groundbreaking visual AIDE (Artificial Intelligence Development Environment) for no-code data pipelines and multimodal agents. Magick can connect to other services and comes with nodes and templates well-suited for intelligent agents, chatbots, complex reasoning systems and realistic characters.