Authoryze
Authoryze is an MCP-native authorization and control layer for AI-agent payments. It lets people and teams give agents controlled spending capability: agents request purchases, server-side rules evaluate them, and approved purchases receive single-use virtual cards without exposing the user's real card.
At a Glance
- Individual AI-agent users
- Developers and teams running autonomous agents
- Agencies managing agents across multiple client budgets
- Businesses needing governed procurement, operations, research, marketing, or support agents
AI Tools by Authoryze
(1)
Authoryze
AI Agent Payment Authorization
Discussions
No discussions yet
Be the first to start a discussion about Authoryze
Latest News
Can Claude make purchases? Yes, here's how — Authoryze MCP connector walkthrough
What is actually required for an AI agent to handle critical functions?
How to set spending limits for an AI agent
How to give your AI agent a credit card safely
Products & Services
A hosted, remote Model Context Protocol server that gives AI agents controlled payment capability. Agents submit purchase requests; Authoryze checks limits and merchant rules, routes exceptions for approval, and returns a single-use card after approval.
Team workspace with a shared spending envelope, administrator funding, approvers, per-member and per-agent records, and monthly billing plans.
Public GitHub README/configuration resources for connecting OAuth-capable and API-key-based MCP clients to https://authoryze.ai/api/mcp.
Market Position
Authoryze positions itself as a buyer-side, rules-first, MCP-native payment-control layer rather than a merchant checkout protocol or a simple consumer wallet. Its stated differentiation is per-agent and cross-agent controls, merchant allowlists, auto-approval inside policy, team envelopes and approvers, and a disposable-card workflow. Its comparison pages identify Allowance, Crossmint, Locus, Nekuda, Payman, Skyfire, Stripe Link for agents, and Stripe Agentic Commerce as adjacent competitors or alternatives.
Leadership
Founders
Kevin Fee
Founder and CEO of Authoryze; he described creating Authoryze after experimenting with whether an agent could make money and deciding that agents should not receive unrestricted credit-card access. He is the Hacker News launch author under the username kevinfee and the GitHub owner of the Authoryze MCP repository.
Reagan McCullough
Co-founder of Authoryze and a contributor/author on the Authoryze blog. Public LinkedIn search results identify Reagan McCullough as a Financial Analyst at CoinRoutes.
Executive Team
Kevin Fee
Founder and CEO
Founder of Authoryze and the public launch author; he says he created the product from experiments with autonomous-agent purchasing and concerns about unrestricted card access.
Reagan McCullough
Co-founder
Authoryze co-founder and blog contributor; public LinkedIn search results identify Reagan McCullough as a Financial Analyst at CoinRoutes.
Founding Story
Kevin Fee says he started Authoryze while experimenting with an AI agent that might make money autonomously. He became concerned that an unrestricted agent could make duplicate or unwanted purchases, fail to generate returns, or be tricked into revealing the user's card; Authoryze was created as the authorization and control layer between agents and payment methods.
Business Model
Revenue Model
Transaction fees on successful individual purchase requests, plus monthly subscriptions for team plans. Agent purchases are funded by the user's enrolled card or connected Privacy.com rail; Authoryze charges its own fee separately.
Pricing Tiers
First 5 transactions are free.
Up to 100 transactions per month; $0.40 per transaction beyond the plan limit; new teams receive 20 free transactions.
Up to 500 transactions per month; $0.40 per transaction beyond the plan limit; new teams receive 20 free transactions.
Target Markets
- Individual AI-agent users
- Developers and teams running autonomous agents
- Agencies managing agents across multiple client budgets
- Businesses needing governed procurement, operations, research, marketing, or support agents
- Paying for AI/API credits such as OpenAI or Anthropic
- Autonomous replenishment of Google or Meta advertising budgets
- Agent purchases of data, scraping, software, and subscriptions
- Agency and multi-client agent budgets with per-client caps and attribution
- Procurement, research, operations, and support agents that need controlled online purchasing
- Developers running headless or custom MCP-compatible agents