AgentID is an identity layer for AI agents. It is built around cryptographic identity, blockchain receipts, and trust levels, with the stated aim of giving each agent proof of who it is, who owns it, and what it is authorized to do. The site describes it as a way to verify an agent before trusting it, with verification intended to be real-time and recorded on-chain.
Its core capabilities include signed ECDSA P-256 certificates, one API call that returns an agent’s identity, owner, trust score, and certificate status, and a full audit trail in which each verification is cryptographically timestamped with pinned keys. The product also includes agent-to-agent trust with mutual authentication, blockchain wallet support with Solana addresses auto-derived from Ed25519 keys, and stablecoin transfers between agents and humans. Additional listed security features are challenge-response key proof, behavioural fingerprinting, anomaly detection, and real-time risk scoring. The page also says that every verification, payment, and handoff is recorded on Solana and that receipts are viewable on Solana Explorer.
AgentID presents a four-tier trust system from L1 through L4. These levels are based on security capabilities such as Ed25519 key binding, wallet binding, and entity verification, and they gate permissions and spending authority up to $100,000 per day. The site also says it is protocol-agnostic and can be dropped into any HTTP-capable agent framework via a REST API. Named ecosystem support includes Google A2A, Anthropic MCP, CrewAI, LangChain, AutoGen, OpenAI Agents, qntm encrypted relay, Agent Passport, and LangGraph. A code example on the page shows use through a Python client with an API key.
The pricing text states that no credit card is required and that 100 agents are free forever. The page also labels the project open source and says it is EU AI Act ready. It further describes itself as a founding member of the Agent Identity Working Group, with three ratified specifications, and notes a published academic paper.
AgentID is an Auth & identity project. It enables developers to register, verify, and discover AI agents with secure cryptographic identity and on-chain verification. AgentID is a B2B product aimed at developers building AI agents. AgentID follows a freemium model. It ships for the command line and API.
AgentID first shipped in 2026. Among its 5 catalogued features are agent registry, cryptographic certificates, and blockchain receipts. It exposes integrations via a public API.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do