CoinW Research Institute Popular Science: Understand the X402 Protocol in 3 Minutes: The True Starting Point of AI Autonomous Payment

1. Concept Introduction

The X402 protocol is an open payment protocol launched by Coinbase. The "402" in its name comes from the common HTTP status code, meaning "payment required to continue access." The problem it aims to solve is straightforward: enabling AI or software agents to shop, settle, and pay online like humans. This process involves identity verification, data interaction, and API calls, all autonomously completed by AI. We only need to set permissions and spending limits in advance. Technically, X402 serves as the "payment interface" of the internet: the Agent initiates the request; the Facilitator (payment node) verifies and completes the settlement; the Merchant (merchant/developer) returns the result after receiving stablecoins.
In this way, AI can directly use stablecoins like USDC for cross-chain payments without needing accounts or keys. Coupled with the AP2 protocol led by Google—AP2 is responsible for "who can pay," while X402 is responsible for "how to pay"—the combination forms a closed loop of "authorization → execution → settlement" for AI.

2. Track Overview

The recent X402 hype began with the Coinbase Commerce team releasing new documentation, opening up standards and SDK interfaces. Subsequently, more and more projects started building products around it. Currently, the representative projects include:
PayAI
Project Introduction: PayAI is a new project on the Solana chain, focusing on the concept of "AI agents hiring each other to complete tasks." It serves as a platform for AI agents to work and collaborate, allowing different AIs to issue tasks, accept tasks, and settle rewards via blockchain. The product has high usage and penetration, with PayAI processing over 14% of the entire X402 ecosystem's transaction volume. PayAI is also the first payment node on Solana to implement the X402 protocol, ranking high in transaction share within the ecosystem.
Token Economic Model: The total supply of PayAI tokens is 1 billion, adopting a 100% initial circulation structure, with no private placements, no lock-ups, and no team reserves. At launch, the PayAI team purchased 20% of the tokens at the issuance price and transferred them to the project treasury for operations, community incentives, and promotional collaborations. The treasury tokens are divided into two parts: half for providing liquidity and earning transaction fees; the other half for a 1-year linear release.
Token Functions:

1. Used for fee deductions in task transactions;
2. Enhances agent exposure and ranking in the task market;
3. Used for governance voting and proposal decisions;
4. Serves as a dispute arbitration deposit;
5. Undertakes settlement incentives and node rewards within the X402 payment layer.
   PingAI
   Project Introduction: PingAI is a routing layer node deployed on the Base chain, directly connected to Coinbase Commerce. Tokens are minted through the X402 protocol, using a "hardcore minting" method reminiscent of the early inscription craze—without a polished front-end, purely on-chain operations, leaning towards a geek style. The project was initially positioned as a pure meme token but has recently transformed into a Launchpad platform token, reducing reliance on emotional consensus. Meanwhile, the team plans to expand asset issuance narratives in the X402 track, covering more application scenarios like GameFi and SocialFi.
   Token Economic Model: The PingAI token is minted on the Base chain using a "pay-as-you-go" model. Early users can pay $1 to receive 5,000 $PING tokens, with a large initial supply and a community distribution model due to no lock-ups or pre-sale phases, resulting in extreme supply dispersion and significant price volatility.
   Token Functions:
6. As the first issued token of the X402 protocol, it symbolizes the ecosystem and serves as a consensus carrier;
7. Used for community activities, task participation, and ecosystem incentives;
8. Future plans include node staking and liquidity governance;
9. It has meme attributes, representing the cultural consensus of the X402 ecosystem.
   GoKite AI
   Project Introduction: GoKite AI is building an AI payment blockchain where autonomous AI agents can operate under verifiable identities and programmable governance, with native access to stablecoin payments. As of the end of October 2025, the GoKite AI network has processed over 600 million AI calls, with approximately 13.6 million users onboarded. Combining AI agents, autonomous execution, and on-chain payments, it is a rare Layer 1 project in the AI + payment track.
   Funding and Team: GoKite AI gained early institutional favor, raising $18 million in Series A funding led by PayPal Ventures and General Catalyst. GoKite AI was formerly known as Zettablock, with team members from Uber, Databricks, Salesforce, NEAR, etc., collaborating with platforms like PayPal and Shopify to build the "foundation layer of the agent internet."
   Token Economic Model: The total supply of GoKite AI tokens is 10 billion KITE. The distribution structure is as follows: ecosystem and community: 48%; Kite modules and tool layer: 20%; team, advisors, and early contributors: 20%; investors: 12%. The token release plan is oriented towards long-term incentives, with team and investor shares unlocked in phases.
   Token Functions:
10. Used for payments and settlements of AI agent tasks;
11. Serves as the foundational token for governance and reputation rating;
12. Participates in protocol upgrades and proposal voting;
13. Used for testnet and task incentives;
14. Future support for module integration and cross-chain task settlement functions.
    Questflow
    Project Introduction: Questflow is not a single AI tool but a platform that enables multiple AI entities to collaborate, akin to an "AI employee scheduling system." Future AI will not just be tools but a new form of labor. Every team may have its own AI assistants and AI employees, and Questflow is building this "AI collaboration hub." It is like the App Store of the AI era.
    Funding and Partnerships: It recently completed a $6.5 million funding round, with investors including Coinbase, Circle, Animoca, HashKey, Aptos, and Ethereum co-founder Joseph Lubin. Questflow has already partnered with Google, Near, Coinbase, and Circle.
    TGE Date: Not yet announced.
    The emergence of projects like GoKite AI and Questflow indicates that capital is shifting from short-term speculation to infrastructure-level AI payment layouts. Perhaps in the future, there will be more institutional preferences for the X402 protocol direction. We must also be wary of the "halo effect" of institutions and look beyond appearances to see the essence. Why do top institutions invest in it? Are they focused on real technological barriers or narrative potential? Is the project merely a concept or already being implemented? How is the execution capability of the founding team and the project progress? What is the actual feedback from users and traditional merchants? These may be the areas we should pay attention to as investors.
    Currently, the X402 topic on social platforms has continued to surge in popularity over the past two weeks, with the market discussing whether it could be the "HTTP moment" of the AI economy. Initially, after PayAI's launch, it surged 80 times, and with the recent launch of GoKite AI on Binance, the entire track's hype continues to rise. Notably, beyond the hype, more and more investors and developers are returning to rationality. The three main concerns now are: the number of merchant integrations, the number of real payment transactions, and developer activity. If these three can continue to grow, X402 can be considered to have truly moved from narrative to implementation.

3. Background Source

From Coinbase's initiation to the collaborative construction of an open payment standard, X402 is not solely dominated by Coinbase; it is more like an "industry co-built open standard." Teams such as Base, Cloudflare, Circle, LangChain, and Autonolas are participating in different directions to improve it. Initially, the concept of X402 was proposed by the Coinbase Commerce team, but it quickly evolved into an open standard co-built by multiple parties. The person in charge, Sid Coelho-Prabhu, is responsible for payment operations at Coinbase, and he led the transformation of this protocol from a "concept" into a system that can run on-chain. The underlying performance is supported by the Base chain, with settlements completed by Circle's USDC, and the AI agent framework integrates LangChain and Autonolas. Thanks to the collaborative efforts of these parties, X402 has transitioned from an experimental standard to a truly usable open protocol. At the same time, the division of labor within the entire ecosystem has gradually become clear, as some are working on SDKs, some are enhancing payment interfaces, and others are improving identity verification and stablecoin settlement layers. It is this open collaboration that has allowed X402 to move from the experimental stage to an industry-level application framework.
In simple terms, Coinbase is the igniter but not the only driver. The entire track is exploring "how AI learns to pay." The ultimate goal is to make X402 the industry universal language for AI payments, rather than a proprietary protocol of a single company.
For the Web3 community, this represents a long-lost spirit of co-construction. It is reminiscent of the early BTC/ETH forums, where developers from around the world contributed code, filled gaps, and iterated together on GitHub. This may be the most primitive consensus in our crypto world—focusing on doing one thing well, taking root and sprouting, and whether it changes the world depends on everyone's belief. After all, isn't the essence of blockchain such a spirit of "open collaboration rather than noisy speculation"?

4. Outlook Prediction

Overall, the embryonic form of the X402 ecosystem has taken shape. PING has secured the first consensus, PayAI has successfully enabled cross-chain payments, and GoKite has brought AI execution onto the chain. The three are like three well-fitted baseboards, pushing the concept of "AI can pay for itself" into reality. Together with AP2, X402 forms a new payment infrastructure—a system that truly endows machines with economic identities. The standards are still evolving, and the number of implemented projects is limited, but the trend is set: the first expenditure of the AI economy is likely to start from X402.
Just as HTTP facilitated the free flow of information, X402 may be the channel that truly enables the flow of "value." RWA brings real assets onto the chain, while X402 allows these on-chain assets to flow back into real consumption scenarios, forming a new value closed loop. Perhaps one day in the future, what truly brings crypto into the mainstream is not ETFs, but merchants like Starbucks, McDonald's, and Amazon starting to directly accept USDC, USDT, or even USD1. At that time, crypto payments will no longer be a concept but will enter everyone's life. If 2023 is the year of the explosion of AI content, then 2025 is likely to be the year of AI payment takeoff, and X402 could very well be the starting point of that transformation.