Original | Odaily Planet Daily (@OdailyChina)

Author｜jk

The Solana x402 Hackathon, which lasted for two weeks, successfully concluded in November, and the organizers officially announced the winners of the main competition track on November 25. This remote hackathon attracted enthusiastic participation from developers worldwide, ultimately receiving over 400 project submissions. The previously popular AI payment protocol x402, developed by Coinbase, aims to enable AI programs to autonomously complete online payments like humans. The vision is for your AI assistant not only to help you search for information but also to pay for data and subscribe to services on its own, all automatically completed on the blockchain.

The hackathon established five competition categories, with a maximum prize of $20,000 for each category. Now, let Odaily Planet Daily take you through the innovations of these five winning projects.

Intelligence Cubed (i³): Trading AI Models Like Stocks

Intelligence Cubed has created an interesting platform that can be understood as "Taobao + stock market for AI models." On this platform, AI models can not only be used but also bought, sold, and invested in.

Imagine this scenario: you are a developer of an AI model and have spent a lot of time training a powerful image recognition model. In the traditional model, you might need to set up your own server, handle payments, and manage users. But on the i³ platform, you only need to upload the model and set the price for each call (e.g., $0.01), and the platform will handle everything automatically.

Even more interestingly, i³ introduces the concept of "model tokenization." Developers can sell ownership of the model in multiple shares through IMO (Initial Model Offering, similar to an IPO). After investors purchase model tokens, every time someone uses the model and pays a fee, token holders receive a proportional share of the revenue. If someone creates an improved version based on your model, your original model can automatically receive "royalties." The project also proposes the concept of "open-source threshold," where the model will automatically become open-source when more than 51% of its ownership is held by the public, accelerating adoption and recreation.

In terms of technical implementation, i³ deeply integrates the x402 payment protocol. Each time a user wants to call an AI model, the system first generates a payment request showing how much USDC needs to be paid. After the user confirms the payment through the Phantom wallet, the transaction is verified on the Solana blockchain, and the entire process takes only a few seconds. The AI model will only start working and return results after payment confirmation. The platform also provides a visual workflow editor, allowing users to connect multiple AI models like building blocks to create complex processing flows, with clear costs at each step.

PlaiPin (Solana ESP32 Native x402): Teaching IoT Devices to Spend Money

What PlaiPin is doing sounds a bit sci-fi: they enable a microchip (ESP32) that costs only a few dollars to manage its own wallet and make payments. What does this mean?

Imagine you have a smart temperature sensor that collects data every day. In the traditional model, this sensor needs to send data to a cloud server, where humans decide whether to sell the data. But with this technology, the sensor itself can become an independent "merchant": it can determine when the data is valuable, contact buyers, collect payments, and store the money in its own blockchain wallet.

For example, your smart refrigerator detects that it needs to call an AI service to optimize the temperature control algorithm; it can autonomously pay $0.001 to purchase this service without any human intervention. Or your robotic vacuum cleaner encounters complex terrain while cleaning and needs to purchase a call to a premium navigation algorithm; it can also complete the payment on its own.

Technically, the breakthrough of this project lies in embedding a complete blockchain wallet and payment capability into a small chip. The ESP32 chip stores its own keys (like a bank card password) and can perform encrypted signatures to prove "this payment is indeed what I want to make." The entire payment process takes about 2-4 seconds: the device identifies the need for a paid service, automatically parses the price and payment address, signs the transaction internally, submits it to the blockchain network through a facilitator (understood as a payment channel), and finally obtains the service. The key point is that the user's wallet private key never leaves the chip, ensuring security.

The project code has been tested on real hardware, and developers have provided detailed installation guides, allowing anyone to purchase a set of hardware for a few dozen dollars to try it out. This opens up a new business model for IoT devices: making them "electronic life forms" that can actively participate in economic activities.

x402 Shopify Commerce: Enabling Taobao Stores to Serve AI Customers in 2 Minutes

If the previous projects were more technical, the x402 Shopify Commerce project is very down-to-earth. It aims to solve the problem of how ordinary online stores can serve AI customers.

Current online stores are designed for humans: they have images, shopping carts, and checkout buttons. But AI programs "cannot understand" these. This project is like installing an "AI-specific channel" for online stores: store owners only need to do three things—first, paste the URL and authorization code of their Shopify store (30 seconds); second, select which products are allowed for AI purchase (60 seconds); third, open the monitoring panel to view orders placed by AI (30 seconds). The entire process requires not a single line of code.

Once set up, AI programs can shop like humans. For example, an AI assistant from a company receives a task to "order 100 pens for the office"; it will automatically search for your store, check the product catalog, select suitable products, calculate the total price, and then pay with USDC. The entire process follows the standard x402 protocol: the AI initiates a purchase request, your store automatically tells the AI "you need to pay X dollars USDC to this address," the AI completes the transfer, and once the store verifies the payment, it automatically creates an order, which will appear in your Shopify backend like a regular order, and you can ship it as usual.

This project cleverly combines two open standards: MCP (Model Context Protocol) allows AI to "understand" what products your store has, and x402 standardizes and automates the payment process. More importantly, because it uses direct blockchain transfers, store owners do not need to pay credit card fees (usually 3-5%), and the funds can arrive in seconds.

For early-stage AI startups, this means they can allow their AI products to purchase resources directly from suppliers without human approval or pre-funding. For e-commerce sellers, this opens up a new customer base—AI agents that autonomously procure on behalf of companies or individuals.

Amiko Marketplace: Establishing Credit Profiles for AI

As AI programs begin to spend money on services, a question arises: how do I know if this AI is reliable? Will it run away after making a payment? How good is the quality of the services it provides? Amiko Marketplace aims to solve this problem by establishing a "credit profile" for each AI on the blockchain.

The operation of this system is quite clever. Each time an AI program receives its first payment, the system automatically creates an identity profile for it, recording its wallet address and basic information. Every time the AI completes a task and receives payment, the system creates a permanent work record, including who the client is, how much was paid, transaction hash, and other information. After using the service, clients can rate the AI (1-5 stars) and leave feedback.

The most interesting aspect is its scoring mechanism: it does not simply take the average score but "weights it by payment amount." Suppose an AI receives 5 stars for a $100 transaction and 3 stars for a $10 transaction; its overall score will be closer to 5 stars because the evaluation weight of larger transactions is higher. The benefit of this design is to prevent score manipulation—if someone wants to boost their rating through numerous small transactions, the cost will be high, and the effect will be limited.

For example, you develop an AI translation service that initially has no reviews. A client spends $50 using your service, is very satisfied, and gives it 5 stars; your profile now has its first good review and a record of "total transaction amount $50." As more clients use and rate your service, your credit score will increase, and other potential clients will see that you have over 100 positive reviews and a total transaction amount of $10,000, making them more willing to choose your service.

This system also has a "lazy registration" mechanism: new AIs do not need to register in advance; as soon as someone makes a payment to them, the system will automatically create a profile. This lowers the entry barrier, allowing any AI program to immediately start providing services and building a reputation. All work records, evaluations, and scores are permanently stored on the Solana blockchain, where anyone can view and verify them, but no one can tamper with them.

MoneyMQ: Turning Payment Systems into Configuration Files

The final winning project, MoneyMQ, is a developer tool with the philosophy that "payment systems should be as simple as writing configuration files."

In Web2, if you want to add payment functionality to your application, you need to: register with a payment service provider, integrate their SDK, write code to handle various payment statuses, set up a testing environment, handle refunds and disputes… This process can take weeks or even months. MoneyMQ simplifies all of this into "writing a few lines of YAML configuration files on your laptop."

Imagine YAML as a product or a set of game rules; it might look something like this:

Product Name: Premium API Access

Price: 0.1 USDC

Billing Method: Per call

You write these rules locally, and MoneyMQ will automatically launch a complete payment environment, including product catalogs, billing logic, test accounts, and more. You can simulate the entire payment process on your own computer: initiate payment requests, verify the x402 protocol, and check fund arrivals. Once testing is successful, you can deploy to the production environment with one click, and all configurations will take effect automatically. MoneyMQ has built-in support for the x402 and MCP protocols. This means that AI programs can not only use your services but also understand your billing rules and even help you optimize pricing strategies. For example, the AI can analyze "how much call volume would increase if the price is lowered from 0.1 USDC to 0.08 USDC" and then suggest you adjust the price.

The "embedded yield" feature planned for the project is also very creative: the balance in your account will not sit idle but will automatically participate in DeFi (decentralized finance) yield strategies. For instance, if you earn 1000 USDC this month, before you decide to withdraw, this money will automatically earn an annualized yield of 4-5%. This can be a significant additional income for businesses with substantial cash flow.

MoneyMQ has already provided a Homebrew installation package for macOS, allowing developers to install it with a single command.

In Conclusion

Of course, these projects are still in their early stages, but the possibilities they showcase are already exciting enough. For ordinary users, these technologies may still seem a bit distant. But imagine this: perhaps in the near future, your smart home system will purchase weather forecast services on its own to decide whether to water the plants, your dashcam will sell the traffic information it captures to mapping companies, and your health monitoring wristband will pay to use the latest AI diagnostic models… When AI can autonomously handle these small payments, our digital lives may become smarter and more convenient.

The organizers stated that the winners of the partner track will be announced next week.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。