--- license: mit --- # Benchmark Agent Meta and Traffic Dataset in AI Agent Marketplace | AI Agent Directory | AI Agent Index from DeepNLP This dataset is collected from AI Agent Marketplace Index and Directory at http://www.deepnlp.org, which contains AI Agents's meta information such as agent's name, website, description, as well as the monthly updated Web performance metrics, including Google,Bing average search ranking positions, Github Stars, Arxiv References, etc. The dataset is helpful for AI researchers and practitioners to do analysis, write reports, monitor the trends and track the exponentially growth (though flat) of number of AI Agents in all fields. To add and list your agent to the Agent Index, please visit [AI Agent Directory](http://www.deepnlp.org/store/ai-agent), use the the workspace [http://www.deepnlp.org/workspace/my_ai_services](http://www.deepnlp.org/workspace/my_ai_services) or using python API To Search all agents, please use AI Agent Search Engine [Agent Search Engine](http://www.deepnlp.org/search/agent) [Github AI Agent Marketplace](https://github.com/AI-Agent-Hub/AI-Agent-Marketplace) [Pypi AI Agent Marketplace](https://pypi.org/project/ai-agent-marketplace/) ![AI Agent Marketplace Directory Search](https://raw.githubusercontent.com/AI-Agent-Hub/AI-Agent-Marketplace/refs/heads/main/docs/image_email_writing_agents.jpg) # Data Samples ## [github com](https://github.com/TheAgentCompany/TheAgentCompany)
### website https://github.com/TheAgentCompany/TheAgentCompany ### Agent Page http://www.deepnlp.org/store/ai-agent/benchmark/pub-github-com/github-com ### description ### category Benchmark ### tags
## [openai com](https://openai.com/index/mle-bench/)
### website https://openai.com/index/mle-bench/ ### Agent Page http://www.deepnlp.org/store/ai-agent/benchmark/pub-openai-com/openai-com ### description OpenAI's Deep Research is an advanced AI agent integrated into ChatGPT, designed to autonomously conduct comprehensive, multi-step research tasks on the internet.

Introduction

Deep Research l ### category Benchmark ### tags
## [aimultiple](https://research.aimultiple.com/ai-coding-benchmark/)
### website https://research.aimultiple.com/ai-coding-benchmark/ ### Agent Page http://www.deepnlp.org/store/ai-agent/benchmark/pub-aimultiple/aimultiple ### description AI coding benchmarks are standardized tests designed to evaluate and compare the performance of artificial intelligence systems in coding tasks. Benchmarks primarily test … ### category Benchmark ### tags
## [convergence ai](https://convergence.ai/introducing-proxy/)
### website https://convergence.ai/introducing-proxy/ ### Agent Page http://www.deepnlp.org/store/ai-agent/benchmark/pub-convergence-ai/convergence-ai ### description Convergence’s Proxy ahead in top agent benchmark, beats OpenAI and Anthropic We are extremely excited to announce the launch of Proxy, our automated AI assistant. ... For example, our Parallel Agents ### category Benchmark ### tags
## [crypto briefing](https://cryptobriefing.com/deepseek-impact-ai-market/)
### website https://cryptobriefing.com/deepseek-impact-ai-market/ ### Agent Page http://www.deepnlp.org/store/ai-agent/benchmark/pub-crypto-briefing/crypto-briefing ### description The AI agent market cap has plunged over $2.5 billion in the past 24 hours, dropping from $12.9 billion to just over $10.2 billion, according to data from cookie.fun.. … ### category Benchmark ### tags
## [hajime ai](https://www.hajime.ai/)
### website https://www.hajime.ai/ ### Agent Page http://www.deepnlp.org/store/ai-agent/benchmark/pub-hajime-ai/hajime-ai ### description Focus your efforts and resources on developing your AI Agents, while we handle the rest and prepare you for a future where agents will work together to innovate. Our platform is designed to offer an e ### category Benchmark ### tags
## [princeton edu](https://agents.cs.princeton.edu/)
### website https://agents.cs.princeton.edu/ ### Agent Page http://www.deepnlp.org/store/ai-agent/benchmark/pub-princeton-edu/princeton-edu ### description AI agents are an exciting new research direction, and benchmarks are crucial for driving progress. However, current agent benchmarks and evaluation practices reveals several shortcomings that hinder t ### category Benchmark ### tags
## [aimultiple com](https://research.aimultiple.com/ai-agents/)
### website https://research.aimultiple.com/ai-agents/ ### Agent Page http://www.deepnlp.org/store/ai-agent/benchmark/pub-aimultiple-com/aimultiple-com ### description AI agents are autonomous software systems that plan and act to achieve given tasks or goals. They are equipped with aggregated knowledge and experience of human experts and access to relevant data. We ### category Benchmark ### tags
## [agentclinic github io](https://agentclinic.github.io/)
### website https://agentclinic.github.io/ ### Agent Page http://www.deepnlp.org/store/ai-agent/benchmark/pub-agentclinic-github-io/agentclinic-github-io ### description @misc{schmidgall2024agentclinic, title={AgentClinic: a multimodal agent benchmark to evaluate AI in simulated clinical environments}, author={Samuel Schmidgall and Rojin Ziaei and Carl Harris and Edua ### category Benchmark ### tags
## [langbase com](https://langbase.com/state-of-ai-agents)
### website https://langbase.com/state-of-ai-agents ### Agent Page http://www.deepnlp.org/store/ai-agent/benchmark/pub-langbase-com/langbase-com ### description → Dismiss ⌘ NEW Agent Architectures Composable AI Email Agents ⌘ Step #0 Step #0 # mkdir ai-email-agent && cd ai-email-agent mkdir ai-email-agent && cd ai-email-agent Copy Copied! npm init -y && ### category Benchmark ### tags
## [mit edu](https://aiagentindex.mit.edu/)
### website https://aiagentindex.mit.edu/ ### Agent Page http://www.deepnlp.org/store/ai-agent/benchmark/pub-mit-edu/mit-edu ### description To fill this gap, we introduce The AI Agent Index, the first public database to document information about currently deployed agentic AI systems. How did we collect the data? From August 2024 to Janua ### category Benchmark ### tags
## [digital information world](https://www.digitalinformationworld.com/2024/06/new-benchmark-shows-ai-agents.html)
### website https://www.digitalinformationworld.com/2024/06/new-benchmark-shows-ai-agents.html ### Agent Page http://www.deepnlp.org/store/ai-agent/benchmark/pub-digital-information-world/digital-information-world ### description A customer experience AI startup, Sierra, has developed a new benchmark that helps in evaluating the performance of AI chatbot agents. The benchmark is named TAU-bench and is evaluated by having conve ### category Benchmark ### tags
## [Unity ML-Agents](https://github.com/Unity-Technologies/ml-agents/blob/develop/docs/ML-Agents-Overview.md)
### website https://github.com/Unity-Technologies/ml-agents/blob/develop/docs/ML-Agents-Overview.md ### Agent Page http://www.deepnlp.org/store/ai-agent/benchmark/pub-unity-ml-agents/unity-ml-agents ### description Unity Machine Learning Agents Toolkit enables games and simulations to serve as environments for training intelligent agents using deep reinforcement learning and imitation learning, leveraging Unity' ### category BENCHMARK ### tags
## [Webots](https://github.com/cyberbotics/webots)
### website https://github.com/cyberbotics/webots ### Agent Page http://www.deepnlp.org/store/ai-agent/benchmark/pub-webots/webots ### description Webots provides a complete development environment to model, program and simulate robots, vehicles and mechanical systems. ### category BENCHMARK ### tags
## [Jericho](https://github.com/microsoft/jericho)
### website https://github.com/microsoft/jericho ### Agent Page http://www.deepnlp.org/store/ai-agent/benchmark/pub-jericho/jericho ### description Jericho is a lightweight python-based interface connecting learning agents with interactive fiction games, which focuses on interactive fiction games, enabling research in natural language understandi ### category BENCHMARK ### tags
## [OpenAI Gym](https://github.com/openai/gym)
### website https://github.com/openai/gym ### Agent Page http://www.deepnlp.org/store/ai-agent/benchmark/pub-openai-gym/openai-gym ### description OpenAI Gym is A standard toolkit for developing and comparing reinforcement learning (RL) algorithms. It offers a variety of tasks like Atari games and control problems. ### category BENCHMARK ### tags
## [Habitat](https://aihabitat.org/)
### website https://aihabitat.org/ ### Agent Page http://www.deepnlp.org/store/ai-agent/benchmark/pub-habitat/habitat ### description Habitat: A Platform for Embodied AI Research is the original paper published on ICCV. Habitat enables training embodied agents (virtual robots) in highly efficient photorealistic 3D simul ### category BENCHMARK ### tags
## [PettingZoo](https://pettingzoo.farama.org)
### website https://pettingzoo.farama.org ### Agent Page http://www.deepnlp.org/store/ai-agent/benchmark/pub-pettingzoo/pettingzoo ### description PettingZoo is a python interface capable of general multi-agent reinforcement learning (MARL) problems. PettingZoo includes a wide variety of reference environments, helpful utilities, an ### category BENCHMARK ### tags
## [SOTOPIA](https://docs.sotopia.world/)
### website https://docs.sotopia.world/ ### Agent Page http://www.deepnlp.org/store/ai-agent/benchmark/pub-sotopia/sotopia ### description SOTOPIA is the interactive Evaluation for Social Intelligence in Language Agents website github ### category BENCHMARK ### tags
## [AirSim](https://github.com/microsoft/AirSim)
### website https://github.com/microsoft/AirSim ### Agent Page http://www.deepnlp.org/store/ai-agent/benchmark/pub-airsim/airsim ### description AirSim is developped by Microsoft Research as a simulation platform for AI research and experimentation. For example, drone delivery is no longer a sci-fi storyline—it’s a business realit ### category BENCHMARK ### tags