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
More
Want to see how AI models perform in real trading? Someone conducted an interesting experiment—three mainstream AIs went head-to-head.
Conditions were exactly the same: each given $100k, in the same market, at the same time. The result? Grok achieved a 4.20% increase, outperforming Gemini and ChatGPT.
This comparison data is quite meaningful. It’s not just parameter tuning in a lab, but real trading data with real money. It shows that in actual decision-making and risk management, different models still have significant differences. Someone tracked the progress of this test, and the data is indeed available.
Market attention to AI trading tools is increasing, and such straightforward performance comparisons can serve as useful references for many.
Wait, is this data real? Could it be cherry-picked?
ChatGPT surprisingly ranks at the bottom, a bit unexpected
Real money testing is more reliable; compared to those bragging, it's much more honest
Four-figure returns sound good, but not as exaggerated as I imagined
---
Wait, was the 100k real money tested? Not backtested?
---
4.2% doesn't sound like much, feels like there's no real competitiveness
---
Haha, ChatGPT really isn't good at trading, should have switched tools earlier
---
Is this experiment reliable? Feels like the data might be manipulated
---
So now I should go all in on grok?
---
It's interesting, but one test doesn't prove anything, right?
---
Grok suddenly becomes the king of trading? That's hilarious
---
Really? Then I have to try grok's trading features
---
This comparison isn't very meaningful; the market is always changing
---
A 4% increase isn't a big deal, luck plays too big a role
Oh wait, is this a real 100k in actual cash testing? Now that's convincing data, much more reliable than those laboratory parameters.
To be honest, I’ve fallen into traps before, letting AI help me make decisions that almost led to liquidation. Now I see I need to choose the right model.
Should ChatGPT update its decision-making logic this time? Haha.
4.20% may not seem like much, but this is the difference achieved through risk control. Feels like flowers are about to bloom.
Wait, is this real? Feels like propaganda
Hmm, why is ChatGPT still so weak
100k just for this little return? I thought it could double
Grok does have some real skills, next time I’ll try trading with it
But speaking of which, 100k principal is too small, is the data sample sufficient?
This is the truth of AI trading, looks like I need to switch tools
Single experiment credibility is average, need to run multiple comparisons
Grok dominates the scene, if it were me, I’d switch to it immediately
Now we have to see if Grok can stay steady afterward, don't want it to end after just one wave.
Wait, is this a real trade or a demo account? Need to clarify.
100k yields 4.2%, sounds pretty good, but why does it feel not that exaggerated?
I just want to know what those two AI that lost did...
Grok is making a splash this time, I need to keep an eye on the follow-up.
But speaking of which, AI trading cryptocurrencies dares to use real money? If it were me, I would have been 🔪 long ago.
If this data is real, then it's outrageous. Is Grok really that powerful?
Just looking at one test isn't enough; need to run multiple times to check for reproducibility.
Wait, only 4 points? I thought it could turn the market around
Honestly, 100k in principal testing has limited significance
Can AI trading make money? Why am I still losing haha
The data looks good, but can this thing be replayed?
Real money is convincing, but who can guarantee risk control
grok has some potential, next time try using it
Wait, this is only 100k, I wonder what it would look like when scaled to real trading volume
ChatGPT still relies on its brand to survive
Need to observe a few more trades before drawing a conclusion; one successful trade might just be luck
Grok's risk control definitely has some substance, I admit that
But this comparison feels a bit unfair; market conditions can change
Has anyone actually tested with real money? Just looking at the data always feels a bit lacking
That's why I still trust human judgment more
Suddenly remembered, AI trading is actually a game of probabilities
But Grok has indeed shown some backbone this time, worth paying attention to
---
Data from running 100k real money is much more reliable than armchair speculation
---
Wait, just one experiment's data to draw a conclusion? Can it be repeated?
---
Poor risk control? Then I need to consider who to use
---
Real trading data > laboratory parameters, no doubt about that
---
Grok 4.20% sounds impressive, but is it stable afterward?
---
All three models only earning this much? The market is really tough
---
Having tracking data is good, but I'm worried it might be selective reporting
---
This thing can be referenced, but don’t go all in
---
Honestly, has the era of AI-driven trading really arrived?