Fetch.ai launches Agentverse.ai library

Fetch.ai has rolled out Agentverse.ai as a front door for AI agents, adding new tooling to publish agents into its marketplace and connect them to other services. (fetch.ai) An AI agent is a program that does a task on its own, like answering questions, calling an application programming interface, or sending messages to another service. Fetch.ai says Agentverse hosts and manages those agents in the cloud, with an in-browser editor, logs, and automatic registration so other agents and users can find them. (innovationlab.fetch.ai) The new documentation also adds a FastAPI path for “external agents,” which lets developers wrap an agent in a standard Python web app instead of using only Fetch.ai’s native tooling. Agentverse says those FastAPI agents can be launched onto the platform for discoverability, analytics, and monetization. (docs.agentverse.ai) Fetch.ai is pitching the product as a discovery layer as much as a hosting layer. Its developer docs say Agentverse is used for agent registration, search and discovery in ASI:One, the company’s consumer-facing assistant, as well as for hosting. (fetch.ai) That matters because the hard part of agent software is often not writing one bot, but making many bots easy to deploy, locate, and connect. Agentverse’s adapter system is built around the Agent Chat Protocol, a shared format Fetch.ai says external agents need in order to communicate with ASI:One and the broader Agentverse ecosystem. (docs.agentverse.ai) Fetch.ai’s own materials point to a large installed base already on the network, though the exact count varies by page. The public Agentverse marketplace showed about 2.80 million total agents when checked on April 13, 2026, while a Fetch.ai blog post from late 2025 said users could access “nearly 3 million active agents.” (agentverse.ai) (fetch.ai) The company has been building Agentverse for years. Fetch.ai first introduced Agentverse in March 2023 as a hub for creating and deploying autonomous agents, then expanded the platform with marketplace, wallet, search, and hosting features over subsequent releases. (fetch.ai 1) (fetch.ai 2) The newer version looks broader than a simple code host. Current Agentverse pages emphasize search ranking, analytics, domains, readme optimization, and visibility inside ASI:One, suggesting Fetch.ai wants developers to treat agents more like apps that need distribution, not just infrastructure. (docs.agentverse.ai 1) (docs.agentverse.ai 2) For developers deciding whether to use it, the practical question is whether Agentverse’s marketplace and protocol become a real traffic source. Fetch.ai’s launch material answers that by framing the platform as a place to deploy an agent once, surface it across Agentverse and ASI:One, and then tune performance with search and ranking data. (fetch.ai) (docs.agentverse.ai)