Futures
Access hundreds of perpetual contracts
TradFi
Gold
One platform for global traditional assets
Options
Hot
Trade European-style vanilla options
Unified Account
Maximize your capital efficiency
Demo Trading
Introduction to Futures Trading
Learn the basics of futures trading
Futures Events
Join events to earn rewards
Demo Trading
Use virtual funds to practice risk-free trading
Launch
CandyDrop
Collect candies to earn airdrops
Launchpool
Quick staking, earn potential new tokens
HODLer Airdrop
Hold GT and get massive airdrops for free
Launchpad
Be early to the next big token project
Alpha Points
Trade on-chain assets and earn airdrops
Futures Points
Earn futures points and claim airdrop rewards
Beijing Zhongguancun College's Milkyway Intelligent Agent System Tops the FutureX Global AI Prediction Rankings
Mars Finance news, April 2. The Milkyway intelligent agent system, independently developed by the information intelligence team of the Beijing Zhongguancun College, topped the global rankings with a score of 60.9 points in the FutureX evaluation. FutureX is an international dynamic evaluation benchmark designed specifically to test “AI prediction capabilities,” initiated jointly by top academic institutions such as ByteDance, Stanford, Fudan, and Princeton University. Its dataset is published on HuggingFace, using yet-to-be-revealed real events (such as geopolitical games, policy directions, and economic fluctuations) to assess the AI’s predictive performance. In this round of evaluation, Grok4 under Elon Musk’s company scored 25.9 points, while Milkyway’s score exceeded double by more than a bit. The team is led by Associate Professor Shu Xin Zheng from the “AI Core” faculty focusing on large-model directions, and is dedicated to exploring next-generation AI core capabilities. This breakthrough depends not only on the model itself, but also on intensive research at the intelligent agent system layer (harness layer), enabling the system to work across long time horizons, coordinate among multiple agents, and perform self-assessment and iteration corrections. (Wide-Angle Observation)