Every agent in the corpus
A field guide to the systems reshaping how work gets done. Explore the agents and open-source harnesses in the corpus, each classified by deployment mode, inferred topology, home country and build approach. Every specimen card makes the operating model behind the product visible.
A sample of the agents we have classified, specced and sourced
This is a sample rather than a census. The catalogue currently holds 103 named agents, and more arrive every month from government, healthcare, SaaS platforms and finance. The ones here are the entrants with the more interesting or bespoke setups.
Show more
The split worth noting is how they are built. 73 of the 103 are managed and another 11 are low-code, so most of the catalogue is bought rather than operated. Suppliers are building for non-technical users, with copilots, managed infrastructure and low-code options that let a business adopt agents across its processes without an engineering team. That convenience carries a running cost, and cost is what tends to slow a business down as it scales towards working natively with agents.
What are the main types of AI agent?
Broadly four. General assistants that handle open-ended work, coding agents that work inside a repository, platforms you build your own agents on, and vertical agents built for one industry such as legal, health or finance. Plenty of products sit in more than one group, because the type follows what a system does rather than how it is sold.
How many AI agents are there?
Nobody has a full count. New ones ship weekly and many are features inside larger products rather than standalone tools. This index tracks just over a hundred named agents that have public documentation to check against.
What is an AI coding agent?
An agent that works inside a codebase rather than a chat window. It reads files, edits code, runs tests and can open a pull request, taking several steps before handing back. Cursor, GitHub Copilot and Claude Code are the widely used examples.
What is a vertical AI agent?
One built for a single industry instead of general use. Legal, health and finance agents carry domain knowledge, the document formats that sector uses, and the audit trail regulated work needs. They are newer than general assistants and growing faster.
Where are most AI agent companies based?
Overwhelmingly the United States. Of the 103 agents indexed here, 89 are headquartered there, with the rest spread thinly across the UK, Canada, France, Australia and China. That reflects where the funding and distribution sit rather than where the engineering talent is.