Interpretation of the Five Winning Projects from Solana's Latest x402 Hackathon

Original Title: "Solana x402 Hackathon Concludes: Five Innovative Projects Stand Out"

Original Author: jk, Odaily

The two-week-long Solana x402 Hackathon successfully concluded in November, with the organizers officially announcing the list of main track winners on November 25. This remote hackathon attracted enthusiastic participation from developers worldwide, ultimately receiving over 400 project submissions. Previously, the popular AI payment protocol x402 was developed by Coinbase as an internet-native payment protocol. Its goal is to enable AI programs to autonomously complete online payments just 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 executed on the blockchain.

This hackathon set up five competition categories, each with a top prize of $20,000. Now, let Odaily walk you through the five winning projects and their innovations.

Intelligence Cubed (i³): Trading AI Models Like Stocks

Intelligence Cubed has built a fascinating platform that can be understood as a "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 an AI model developer who has spent a lot of time training a powerful image recognition model. In the traditional model, you might need to set up servers, handle payments, and manage users yourself. But on the i³ platform, you only need to upload your model and set the price per call (for example, $0.01), and the platform will automatically handle everything else.

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

Technically, i³ deeply integrates the x402 payment protocol. Every time a user wants to call an AI model, the system generates a payment request showing how much USDC needs to be paid. After the user confirms the payment via the Phantom wallet, the transaction is validated on the Solana blockchain, and the entire process takes only a few seconds. Only after payment is confirmed does the AI model begin working and return results. The platform also provides a visual workflow editor, allowing users to chain multiple AI models together like building blocks to create complex processing flows, with clear fees for each step.

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

What PlaiPin does sounds a bit sci-fi: they enable a microchip (ESP32) costing just a few dollars to manage its own wallet and make payments autonomously. What does this mean?

Imagine you have a smart temperature sensor that collects data daily. Traditionally, this sensor would send data to a cloud server, and a human would decide whether to sell the data. But with this technology, the sensor itself can become an independent "merchant": it can judge when its data is valuable, contact buyers on its own, receive payments, and store the money in its own blockchain wallet.

For example, your smart fridge detects the need to call an AI service to optimize its temperature control algorithm. It can autonomously pay $0.001 to purchase the service, with no human intervention required. Or your robot vacuum encounters complex terrain while cleaning and needs to purchase an advanced navigation algorithm call—it can also complete the payment on its own.

Technically, the breakthrough of this project is fitting a complete blockchain wallet and payment capability into a small chip. The ESP32 chip stores its own keys (like a bank card PIN) and can perform cryptographic signatures to prove "this payment is indeed authorized by me." The entire payment process takes about 2-4 seconds: the device detects the need for a paid service, automatically parses the price and recipient address, signs the transaction internally, submits it to the blockchain network via a facilitator (think of it as a payment channel), and finally receives the service. Crucially, the user's wallet private key never leaves the chip, ensuring security.

The project code has already been tested on real hardware, and the developers have provided detailed installation guides. Anyone can try it out by purchasing a hardware kit for a few dozen dollars. This opens up a brand-new business model for IoT devices: enabling devices to actively participate in economic activities as "electronic lifeforms."

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

If the previous projects are more technical, the x402 Shopify Commerce project is very practical. The problem it aims to solve is: how can ordinary online stores serve AI customers?

Current online stores are designed for humans: there are images, shopping carts, and checkout buttons. But AI programs "can't understand" these. This project is like installing an "AI-only channel" for online stores: store owners only need to do three things—first, paste their Shopify store URL and authorization code (30 seconds); second, select which products are available for AI purchase (60 seconds); third, open the monitoring dashboard to view AI orders (30 seconds). The whole process requires no coding.

Once set up, AI programs can shop just like humans. For example, a company's AI assistant receives the task "order 100 pens for the office," and it will automatically search your store, browse the product catalog, select suitable items, calculate the total price, and pay with USDC. The entire process follows the standard x402 protocol: the AI initiates a purchase request, your store automatically tells the AI "pay X USDC to this address," the AI completes the transfer, and after the store verifies receipt, it automatically creates an order, which appears in your Shopify backend like any regular order, and you ship as usual.

This project cleverly combines two open standards: MCP (Model Context Protocol) allows AI to "understand" what products your store offers, and x402 standardizes and automates the payment process. More importantly, because blockchain direct transfers are used, store owners don't have to pay credit card fees (usually 3-5%), and funds are received within seconds.

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

Amiko Marketplace: Building Credit Profiles for AI

When AI programs start spending money to buy services autonomously, a question arises: how do I know if this AI is trustworthy? Will it run away after paying? Is the quality of its services good? Amiko Marketplace aims to solve this problem by establishing a "credit profile" for each AI on the blockchain.

This system operates very ingeniously. Whenever an AI program receives its first payment, the system automatically creates an identity profile for it, recording its wallet address and basic information. Each time the AI completes a task and receives payment, the system creates a permanent work record, including the client, payment amount, transaction hash, and more. After using the service, clients can rate the AI (1-5 stars) and leave a review.

The most interesting part is its scoring mechanism: instead of a simple average, it uses "payment amount weighting." 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, as higher-value transactions carry more weight. This design helps prevent score manipulation—if someone tries to boost ratings with many small transactions, the cost is high and the effect limited.

For example: you develop an AI translation service with no initial reviews. A client spends $50 on your service, is satisfied, and gives 5 stars. Your profile now has its first positive review and a "total transaction amount of $50." As more clients use and rate your service, your credit score increases. When other potential clients see you have over 100 positive reviews and a total transaction amount of $10,000, they are naturally more likely to choose your service.

The system also features a "lazy registration" mechanism: new AIs don't need to register in advance—once someone pays them, the system automatically creates a profile. This lowers the entry barrier, allowing any AI program to immediately start offering services and building a reputation. All work records, reviews, and scores are permanently stored on the Solana blockchain, accessible and verifiable by anyone, but tamper-proof.

MoneyMQ: Turning Payment Systems into Configuration Files

The final winning project, MoneyMQ, is a developer tool based on 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 a payment service provider account, integrate their SDK, write code to handle various payment statuses, set up a test environment, handle refunds and disputes... this process can take weeks or even months. MoneyMQ simplifies all of this to "writing a few lines of YAML configuration file on your laptop."

Think of YAML as a product, or a set of game rules, and it might look like this:

Product Name: Advanced API Access

Price: 0.1 USDC

Billing Method: Per Call

You write these rules locally, and MoneyMQ automatically launches a complete payment environment, including product catalog, 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 receipts. Once testing is complete, you can deploy to production with one click, and all configurations take effect automatically. MoneyMQ has built-in support for x402 and MCP protocols. This means AI programs can not only use your service but also understand your billing rules and even help you optimize pricing strategies. For example, AI can analyze "how much call volume would increase if the price drops from 0.1 USDC to 0.08 USDC," and then suggest price adjustments.

The project's planned "embedded yield" feature is also creative: your account balance won't sit idle but will automatically participate in DeFi (decentralized finance) yield strategies. For example, if you earn 1000 USDC this month, before you decide to withdraw, the funds will automatically earn an annualized yield of 4-5%. For businesses with large cash flows, this can be a considerable extra income.

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

Final Thoughts

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 autonomously purchase weather forecast services to decide whether to water your plants, your dashcam will sell captured road condition data to map companies, and your health monitoring wristband will pay to use the latest AI diagnostic models... When AI can autonomously handle these micro-payments, our digital lives may become even smarter and more convenient.

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