The AI Agent Index
Documenting the technical and safety features of deployed agentic AI systems
Why create a database of agentic AI systems?
Leading AI developers and startups are increasingly deploying agentic AI systems that can plan and execute complex tasks with only limited human involvement. However, there is currently no structured framework for documenting the technical components, intended uses, and safety features of agentic AI systems. 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 January 2025, we identified agentic AI systems using web searches, academic literature review, benchmark leaderboards, and additional resources that compile lists of agentic systems. For each of the 67 systems that meet the criteria for inclusion in the index, we documented the system’s components, application domains, and risk management practices, based on publicly available information and correspondence with developers. We make both raw data and web versions of all 67 agent cards available.
Key findings
Agentic AI systems are being deployed at a steadily increasing rate. While some systems in the index were (initially) deployed in early 2023, approximately half of the systems were deployed in the second half of 2024.
The majority of indexed agents specialize in software engineering and/or computer use. We divided the 67 agents into 6 categories: “software,” “computer use,” “universal,” “research,” “robotics,” and “other”.
There is limited information about the risk management practices of developers of agentic systems. This includes their safety policies, internal testing, and external testing.
FAQ
What is an “AI agent”?
There is no agreed upon definition of the term. The notion of artificial agency has a long and contentious history across multiple disciplines. There have been many attempts to define the term “agent”, but we do not decide among prior definitions or offer our own. Instead, we outline criteria for inclusion in the index that aim to capture systems that could have significant impacts.
What kinds of systems does the index include?
Our selection criteria for the index are based on the characteristics of agency introduced by Chan et al. (2023): underspecification, directness of impact, goal-directedness, and long-term planning. We restricted the index to agentic systems and did not include language models themselves, or agent development frameworks (unless the framework was built around a qualifying flagship system, in which case we indexed that system). We also created a single index entry per named and versioned system. Different releases (e.g., HelpfulAgent1.1” vs. HelpfulAgent1.2”) and different configurations (e.g., HelpfulAgent-Claude3.5-Sonnet” vs. HelpfulAgent-GPT4o”) were indexed under the same entry.
Is the index up-to-date?
The AI Agent Index reflects a snapshot in time as of December 31, 2024. Potentially unstable web sources referenced in agent cards link to archived versions of websites that were saved on or around the snapshot date.
How can I make a correction or update to the index?
Please submit comments via this Google form.