What Is Visa CLI? How Command Line Commerce Works for Developers

Visa CLI is a beta command line tool, built by Visa's Crypto Labs division, that lets developers and their AI agents make programmatic card payments without dealing with API key management or pre-funded accounts. You sign up with GitHub, get approved, and your agent can purchase paid APIs, datasets, and digital services while you code.

For decades, paying for digital services as a developer meant a familiar ritual: sign up for an account, enter your credit card, generate API keys, store them securely, and manage billing dashboards. Each paid service added another set of credentials and another billing relationship.

AI coding agents broke that workflow. When an agent running in your terminal needs to call a paid image generation API or access a proprietary dataset, it cannot pause to fill out a payment form. The agent needs programmatic money movement at the moment of need.

Visa CLI addresses this by attaching a payment credential to the development environment itself. The agent authenticates, the transaction clears through Visa's network, and the developer gets billed. No API keys to rotate. No pre-funded wallets to top up.

This matters because developers now operate in a world where agents generate code, call external services, and chain together workflows autonomously. Payment needs to keep up with that speed.

The broader context reinforces the urgency. Morgan Stanley found that 23% of Americans made purchases using AI in the past month. Salesforce reported that one in five Cyber Week 2025 orders involved an agent. A Worldpay survey found that 59% of shoppers aged 18 to 34 are comfortable with an AI agent browsing and shopping on their behalf. The demand side is moving fast, and developer tools need to follow.

Visa CLI is currently in closed beta, with access gated through a GitHub-based signup at visacli.sh. Based on available documentation, the tool operates through a straightforward flow.

You request access through GitHub authentication. Visa reviews and approves the application (the beta is subject to spending limits). Once approved, you receive install instructions and can start using the CLI from your terminal.

From there, the tool connects your terminal sessions to Visa's payment infrastructure. When your agent or script needs to pay for a service, the CLI handles the transaction. The official site highlights three initial use cases:

Image generation on demand. Your agent calls a paid image gen API, and the CLI pays for it. No pre-funded account required.

Music generation endpoints. Access paid audio and music generation services programmatically. The CLI handles the billing so the agent can focus on creating.

Proprietary datasets. Market data, analytics, and research behind paywalls become accessible without separate subscription management for each provider.

The pattern across all three: remove the billing friction between a developer's agent and the paid service it needs.

Visa CLI did not appear in isolation. It fits within Visa's broader push into what the company calls Visa Intelligent Commerce, a platform designed to let AI agents securely browse, shop, and pay on behalf of users.

The numbers behind this push are concrete. According to Visa's December 2025 announcement, 47% of U.S. shoppers now use AI tools for at least one shopping task. AI-driven traffic to U.S. retail websites surged over 4,700% in the past year. More than 100 partners are building on the Visa Intelligent Commerce platform, with over 30 in the sandbox and 20-plus agent integrations live.

The Intelligent Commerce platform provides tokenized digital credentials, authentication, spending controls, and privacy-aware personalization. These are the same capabilities that underpin what Visa CLI does in the developer context: let a machine-initiated transaction clear through the existing Visa network securely.

For developers building applications that involve any kind of financial orchestration or settlement, this signals a shift in how payments infrastructure will work. The transaction initiator is increasingly software, not a person clicking a button.

One of the harder problems in agentic commerce is trust. If an AI agent walks into a digital storefront and tries to buy something, how does the merchant know the agent is legitimate?

Visa's answer is the Trusted Agent Protocol, released in October 2025 in collaboration with Cloudflare. The protocol uses cryptographic signatures to let merchants verify that an AI agent is authorized, acting on behalf of a real consumer, and operating within defined boundaries.

The protocol includes several components. Agent intent indicators tell the merchant whether the agent is browsing or purchasing. Consumer recognition objects let the agent pass along identifiers like loyalty accounts or device fingerprints so the merchant can connect the interaction to an existing customer relationship. Payment credential handling ensures card data flows through established secure channels.

This matters for Visa CLI users because the same trust framework applies. When your coding agent makes a purchase through the CLI, Visa's infrastructure verifies the agent's identity and ensures the transaction falls within your authorized parameters.

The Trusted Agent Protocol is open source on GitHub, and partners including Stripe, Shopify, Adyen, Microsoft, and Coinbase have provided feedback on the specification. Akamai integrated its bot-detection and behavioral intelligence capabilities into the protocol in late 2025, adding another layer of verification for merchants who need to distinguish legitimate agents from malicious bots.

For developers who want to go deeper than the CLI, Visa also released a Model Context Protocol (MCP) server for Intelligent Commerce. MCP is an open standard that lets AI models and agents connect to external systems through a consistent interface.

Visa's MCP server lets AI agents connect to Visa Intelligent Commerce APIs. This means developers can build agents that enroll cards, initiate purchase instructions, retrieve payment credentials, and manage transactions, all through the MCP protocol that tools like Claude and other LLMs already support.

The Visa MCP repository on GitHub includes Node.js and TypeScript client implementations, authentication patterns, and workflow examples. A companion MCP server provides agents with integration guides and tool definitions to help automate the development process itself.

Alongside the MCP server, Visa launched the Acceptance Agent Toolkit in pilot. Built on the MCP server, the toolkit lets developers and non-technical users trigger payment actions with plain-language commands. You can create invoices, generate payment links, and manage acceptance workflows without writing API integration code.

This toolkit approach reflects a broader trend in developer tooling: meet builders where they work, whether that is a terminal, an IDE, or a chat interface.

Visa CLI is one signal within a much larger market shift. According to Accenture's research on agentic payments, investment in agentic AI applications will reach $1.3 trillion by 2029. The same research found that 87% of financial institutions see trust as the biggest barrier to agentic payment adoption.

At NRF 2026 in January, a live audience survey showed that 75% of retail attendees were implementing or planning agentic commerce initiatives. Stripe launched its Agentic Commerce Suite, Google introduced its Universal Commerce Protocol, and OpenAI's "Buy it in ChatGPT" feature brought agentic purchasing into one of the most widely used AI interfaces.

The common thread across all of this: payments infrastructure built for human-driven checkout flows cannot handle machine-initiated transactions at scale. Traditional checkout assumes a person filling in forms, clicking buttons, and confirming amounts. When the buyer is software, those interaction patterns become bottlenecks.

A Global Payments 2026 Commerce and Payments Trends Report found that 42% of small businesses and 45% of medium-sized firms have already seen AI agents making purchases on behalf of consumers. A Bernstein study estimated that AI agents could increase global e-commerce conversion rates by 1.5 to 2.5 percentage points, generating $240 billion in new revenue.

Whether you are processing a card payment through Visa CLI or routing stablecoin transfers across blockchains, the core challenge is the same. Money needs to move as fast as the software requesting it.

This is why projects focused on programmable payment infrastructure and networked money movement are gaining traction. The next generation of commerce requires payment rails that are API-first, agent-compatible, and capable of clearing transactions in seconds rather than days.

Visa is not alone in building these rails. Stripe launched its Agentic Commerce Protocol (ACP) with partners including Salesforce, Squarespace, and PwC. OpenAI co-developed its own commerce protocol with Stripe for "Buy it in ChatGPT." Google introduced the Universal Commerce Protocol (UCP) at NRF 2026. PayPal partnered with Perplexity on agent-specific checkout flows.

Each approach reflects a different bet on where agentic payments should live in the stack. Visa CLI bets on the terminal. Stripe bets on the checkout layer. OpenAI bets on the chat interface. The developer who understands all three will build more capable agents than the one locked into a single payment pattern.

If you are a developer building AI-powered tools, Visa CLI represents a specific bet: that AI agents will need native payment capabilities, and that embedding those capabilities at the infrastructure level (rather than bolting them on through dashboard-managed API keys) is the right approach.

The timing matters. IDC projects that investment in agentic AI will reach $1.3 trillion by 2029, with agentic applications dominating IT budget expansion. Developers building agents today are making architectural choices that will define how those agents transact for years. Choosing between manual credential management and infrastructure-level payment access is one of those choices. Several practical implications follow.

For agent developers, Visa CLI reduces the setup cost of connecting to paid services. Rather than managing separate billing for each API your agent consumes, the CLI provides a single payment channel.

For merchants and API providers, tools like Visa CLI (along with the Trusted Agent Protocol and MCP server) create a standardized way to accept payments from AI agents. This is relevant if you run a paid API and want to capture revenue from the growing population of coding agents that need your service.

For the broader developer ecosystem, command line commerce validates the idea that payment should be a composable building block in the development stack, not a separate system that requires manual intervention.

The infrastructure powering stablecoin economies faces similar design pressures. Both traditional card networks and onchain payment systems need to solve for the same end state: software-initiated, real-time, programmable money movement.

Getting into the Visa CLI beta involves a few steps:

The beta is limited, and Visa has not published a public timeline for general availability. Given that Visa recently completed hundreds of agent-initiated transactions with partners and projects broader rollouts across Asia-Pacific and Europe in early 2026, the CLI could expand access as the Intelligent Commerce ecosystem matures.

Visa CLI is a beta command line tool built by Visa's Crypto Labs division that lets developers and AI agents make programmatic card payments from the terminal. It removes the need for pre-funded accounts and manual API key management when purchasing paid digital services.

Visa's Crypto Labs division created Visa CLI. It is part of Visa's broader Intelligent Commerce initiative, which provides infrastructure for AI agent-driven commerce.

You can request access at visacli.sh by signing up with your GitHub account. The tool is in closed beta with spending limits and an approval process.

The initial use cases include paying for image generation APIs, music generation endpoints, and proprietary datasets. Any paid digital service that an AI agent needs to access while coding is a potential use case.

Visa CLI itself is a beta tool. You pay for the services and APIs your agent purchases through it, with transactions clearing through Visa's card network. Specific pricing and fee structures have not been publicly detailed for the beta.

Visa's Trusted Agent Protocol is an open framework that uses cryptographic signatures to verify AI agent identity and authorization during commerce interactions. It helps merchants distinguish legitimate agents from bots.

Visa CLI is one tool within the broader Visa Intelligent Commerce platform. Intelligent Commerce provides the infrastructure (tokenization, authentication, spending controls) that makes agent-initiated payments secure and verifiable.

Visa released a Model Context Protocol server that lets AI agents connect to Intelligent Commerce APIs. Developers can use it to build agents that enroll cards, initiate purchases, and manage transactions through a standardized protocol.