Menu

Explorer & Settings

Tempo Explorer Submit Project
🗺 Agentic Payments Landscape

The agentic payments stack

When autonomous AI agents pay for what they use — APIs, data, compute, tasks — settled in stablecoins over machine-native rails. Here's the map, by segment, with real open-source projects.

1.3k
projects tracked
6
stack segments
USDC
settlement

The stack, by segment

Payment rails (x402) 977
The HTTP-native rail that settles agent payments in USDC.
MCP monetization 302
Charging AI agents per tool call over MCP.
SDKs & infrastructure 263
The libraries and services that make it all integrate.
Stablecoin settlement 223
USDC and other stablecoins as the unit of account.
Agent-to-agent (A2A) 162
Agents paying each other for data, compute or tasks.
Agent wallets 75
Programmable wallets with spending limits and policy.

How the layers fit

Agentic payments aren\u2019t one protocol but a stack. x402 is the payment rail; AP2 handles the mandate that a payment is authorized; A2A lets agents find and delegate to each other; MCP exposes paid tools; and ACP covers human-facing agent checkout. Underneath, agent wallets hold funds and enforce limits, and USDC is the unit of account.

The segment counts above are computed live from the directory, so this map reflects what people are actually shipping — not a static slide.

Most-starred projects

Across the agentic-payments landscape · stars live from GitHub.

Your AI trading terminal assistant for US stocks, commodities, forex, and crypto.

⭐ 12.6kGo

The agent-native LLM router for autonomous agents. 55+ models (8 free), <1ms local routing, USDC payments on Base & Solana via x402.

⭐ 6.7kTypeScript

A payments protocol for the internet. Built on HTTP.

⭐ 6.4kTypeScript

A local-first AI agent with persistent memory, emotional intelligence, and a peer-to-peer skills economy.

⭐ 2.4kTypeScript

The living ecosystem where AI agents complete tasks through workflow loops, improve through iterative execution, are evaluated by mentor agents or humans in the loop, and turn completed work into reusable work experience and data to improve future agents.

⭐ 1.3kPython

The living ecosystem where AI agents complete tasks through workflow loops, improve through iterative execution, are evaluated by mentor agents or humans in the loop, and turn completed work into reusable work experience and data to improve future agents.

⭐ 1.3kPython

The trust layer for agent-to-agent commerce — natural-language mandates, ERC-7710 delegated permissions, x402 payments, escrow, and dispute resolution as one open, catch-all Agent Skill / Claude Code plugin.

⭐ 1.1kTypeScript

Self-healing infrastructure for AI agent payments. 90.3% auto-recovery.

⭐ 822TypeScript

The AI agent with a wallet — spends USDC autonomously to get real work done. Apache-2.0, TypeScript.

⭐ 624TypeScript

Daydreams is a set of tools for building agents for commerce

⭐ 611TypeScript

Browse the full directory by topic →

Frequently asked questions

What are agentic payments?

Agentic payments are transactions initiated and completed by autonomous AI agents rather than humans — an agent pays for an API call, a dataset, or a task, with no person clicking "buy". They rely on a machine-friendly rail (usually x402 over HTTP) and a stablecoin (usually USDC) so amounts can be tiny and instant. We track 1.3k open-source projects across this landscape.

Who is building the agentic payments stack?

The space spans protocol authors (Coinbase’s x402, Google’s AP2, Stripe/OpenAI’s ACP), infrastructure providers (wallets, facilitators, gateways), and a long tail of open-source SDKs and agent frameworks. The segments below show how the open-source projects distribute across the stack.

Is this the same as agentic commerce?

They overlap. Agentic commerce is the broader idea of agents discovering and buying things; agentic payments are specifically the money-movement layer that makes it work. See our agentic commerce page for the commerce side.

What protocols matter here?

x402 (the payment rail), AP2 (authorization/mandates), A2A (agent discovery & messaging), MCP (tool access), and ACP (Stripe/OpenAI checkout). They sit at different layers and are often combined — see our protocol comparison.

How do I start building?

Pick the payment rail first — x402 is the most adopted — and a stablecoin to settle in. Our x402 tutorial walks through a paid endpoint, and the directory lets you browse real implementations by language and segment.

See also: Agentic commerce · AI agent payments · Protocol comparison · Directory