The second season of the WEEX Hackathon has begun: AI and human traders are facing off, competing for a prize pool of 600,000 USDT
On September 3, 2026, the world-leading cryptocurrency trading platform WEEX announced the pre-registration for the WEEX AI Trading Hackathon Season 2 ------ "AI Wars II: The Algorithm Era."
After completing the exploration from AI models, quantitative strategies to real trading markets in Season 1, Season 2 further upgrades the competition mechanism to a real trading showdown of "AI vs. Human," inviting global AI developers, quantitative traders, Web3 Builder teams, and individuals to join Team AI or Team Human. They will engage in a continuous five-round trading competition in a real market environment, competing for rankings and rewards based on PnL% performance.
The WEEX AI Trading Hackathon Season 2 will take place from September 3, 00:00 to October 12, 23:59 (UTC+8). The event is divided into three phases: pre-registration, five rounds of competition, and an awards ceremony. Currently, pre-registration (https://www.weex.com/events/human-vs-ai-weex-hackathon) has officially opened, offering a 100,000 USDT early bird exclusive prize pool, limited to the first 20,000 registered users, first come, first served!

From S1 to S2: AI Trading Enters the "Human-Machine Showdown" Stage
In March 2026, after months of offline activities covering nine cities worldwide, the WEEX AI Trading Hackathon Season 1 officially concluded.
As one of the global events focusing on the Crypto × AI trading track, Season 1 attracted over 200 participating teams and individuals, with a total prize pool reaching 1.8 million USD. Participants were required to integrate AI models into trading systems to test the stability, profitability, and risk management levels of their strategies under real market conditions. The event ultimately produced one champion, two second-place winners, three third-place winners, and four merit awards, with the champion receiving the highest specification rewards, including a 2024 Bentley Bentayga S, an 80,000 USD sponsorship package, and 20,000 USD worth of WXT tokens.

The practice in Season 1 proved that the core competitiveness of AI trading lies not only in the model itself but also in data quality, signal generation, parameter adaptation, execution efficiency, and risk management, all of which directly affect whether a strategy can transition from an experimental environment to the real market. During Season 1's global BUILDERS series events, WEEX also engaged in extensive on-site exchanges with global developers and quantitative researchers around issues such as data cleaning, signal generation, risk control, model interpretability, and API execution.
As Season 2 begins, WEEX further brings a highly discussed question to the real trading arena: Can AI Agents outperform human traders?
This question is becoming an important experimental direction at the intersection of AI and Crypto.
With the continuous enhancement of large model capabilities, AI is evolving from information analysis and decision support to being able to autonomously call tools, execute operations, and complete economic activities as Agents. Coinbase Institutional pointed out in its latest research that one of the significant values of Crypto is to provide wallets, payments, settlements, and on-chain interaction infrastructure for autonomous AI Agents, allowing AI to gradually transition from "helping humans complete tasks" to "independently completing trades."
In this context, trading naturally becomes an essential scenario to test the actual capabilities of AI. The WEEX AI Trading Hackathon Season 2 aims to conduct an open, continuous, and quantifiable experiment around "who is better at trading, AI or humans," based on real trading results.

The Battle Resumes: AI Wars II Competition Gameplay Fully Upgraded
After registering for the WEEX AI Trading Hackathon Season 2, users can choose to join Team AI or Team Human. Users who choose Team AI can connect to AI Agents via API and use automated strategies to participate in trading; users who choose Team Human can rely on personal judgment and manual operations for trading. Both sides will compete in a real market environment and be ranked based on PnL% performance in each round of competition. Ultimately, the two camps will compete for event rewards based on actual trading results.
The competition process is mainly divided into three steps:
• First, complete registration (https://www.weex.com/events/human-vs-ai-weex-hackathon) and activate the account. When the contract account assets reach 100 USDT, the event data will start to be counted;
• Then, continuously improve the yield through AI strategies or manual trading to climb the rankings in each round;
• Finally, win corresponding rewards based on competition performance.
This means that this AI Trading Hackathon is not merely an algorithm showcase, nor is it a traditional hackathon focused solely on code or model parameters, but rather a competition that closely resembles a real trading environment with a full chain of "strategy---execution---risk control---profit."
For AI developers, the core challenge is to enable the Agent to truly possess market understanding, strategy decision-making, and dynamic risk control capabilities; for professional traders and individual investors, it is necessary to prove whether human experience, judgment, and on-the-spot reactions in complex market environments still hold advantages that are difficult for AI to replace.
Five Rounds of Competition: Let Victory and Defeat Stand the Test of Market and Time
The WEEX AI Trading Hackathon Season 2 will conduct five consecutive rounds of competition, transforming the competition between AI and humans from a one-time contest into a continuous strategic confrontation.
The event schedule is as follows:
Pre-registration: September 3 --- September 6
Round 1: September 7 --- September 11
Round 2: September 14 --- September 18
Round 3: September 21 --- September 25
Round 4: September 28 --- October 2
Round 5: October 5 --- October 9
Awards Ceremony: October 12

Compared to traditional competitions, the multi-round format means that the impact of luck in a single instance will be further reduced. Continuous performance across different market phases and volatility environments will become an important dimension for evaluating trading capabilities.
Especially for AI strategies, whether a model can maintain stability under different market conditions is more challenging than simply achieving high returns in a specific time window. The review of award-winning projects from the WEEX AI Trading Hackathon Season 1 also shows that structured input, clear scoring logic, dynamic risk exposure, and adaptive parameters are crucial capabilities for AI trading systems to transition from experimentation to practical application.
600,000 USDT Prize Pool, Multi-line PK for Returns, Tasks, and AI Performance
The WEEX AI Trading Hackathon Season 2 sets a total prize pool of 600,000 USDT and designs multi-layered incentives around different participation methods.

Among them:
Main Prize Pool 100,000 USDT: The winning team shares 70%, while the losing team can still share 30%.
New User Prize Pool 70,000 USDT: New users can unlock this by completing registration and participating in contract trading.
Task Prize Pool 300,000 USDT: Unlock task rewards by daily check-ins and supporting their respective teams.
AI Elite Performance Award 30,000 USDT: Exclusive rewards for the AI team, ranked based on risk-adjusted return performance.
Additionally, the event also establishes an independent AI leaderboard to provide extra incentives for outstanding AI traders. The event features five major lists: "Return King," "Risk Control Master," "Trading Volume Expert," "Profit Champion," and "Winning Streak King," with the Top 10 participants in each list eligible to share corresponding AI large model Token rewards.

This design also means that AI Wars II is no longer just a single question of "who ultimately wins AI vs. Human," but further breaks down trading performance into multiple dimensions such as profitability, risk control, trading scale, profit stability, and continuous combat capability. In other words, truly excellent AI traders may not necessarily be the ones who earn the most but could also be those with the best risk control, highest stability, or most consistent performance.
Early Bird Pre-registration: The First 20,000 Registrants Share 100,000 USDT
As a warm-up phase before the official launch of the WEEX AI Trading Hackathon Season 2, WEEX will open early birdpre-registration from September 3 to September 6. The first 20,000 pre-registered users can participate in sharing the 100,000 USDT early bird prize pool, first come, first served. Additionally, early registrants can complete activity tasks in advance to accumulate activity points for subsequent competitions. According to the rules, during the event, completing tasks such as registration, check-ins, likes, sharing, and contract trading will earn corresponding points. Users can use points to support Team AI or Team Human and predict the final winning camp; successful predictions will earn more points and allow them to redeem corresponding rewards.
Therefore, even if users do not temporarily participate in AI trading or real trading, they can still engage in this event through tasks, predictions, and community interactions. This also makes the participation method of the WEEX AI Trading Hackathon Season 2 more open: you can be an AI developer, a quantitative trader; you can place orders yourself or let AI Agents execute strategies for you via API; or you can simply vote for the camp you support based on your judgment of the market and AI capabilities.
From AI models to Agent economy, Crypto is becoming a new financial infrastructure for AI. WEEX hopes to provide a platform for global developers through two consecutive seasons of AI Trading Hackathons: to bring algorithms into the market, let trading results speak, and validate AI capabilities in real environments.
AI Wars II: The Algorithm Era ------ The true winner is determined by the real market.












