# The Final Level is Here #### The Final Level Is Here. Here's Exactly How It Works 4pm UTC. The final level opens. Here's exactly how it works: * Every wallet at Level 5 gets unlocked at the same time * One winner today. First to complete claims 50% of the prize pool. Everyone at Level 5 gets frozen after * Tomorrow: Level 5 resets with a new challenge * Everyone below Level 5 can still advance and compete tomorrow * 25-prompt daily limit returns tomorrow. Today: no cap Fair chances for everyone. Real stakes. If you missed anything, please read the docs or read the previous King Arthur article.\ \ For more real time updates, please join the official telegram channel.\ \ Read below everything you need to know. #### Everyone at Level 5 starts at the same time No head starts. No advantages for whoever got there first. When the final level opens at 4pm UTC, every wallet sitting at Level 5 gets unlocked simultaneously. The playing field is flat. The only thing that matters from that point is what you do next. The first wallet to complete Level 5 claims 50% of the prize pool (currently at 305+ SOLANA), which is growing, paid directly on-chain. Once a winner is confirmed, all remaining Level 5 users get frozen. #### Not at Level 5 yet? You still have a shot Today is not your only window. Tomorrow, Level 5 resets with a new challenge. A different version. A different path to the same prize pool. If you're still climbing, you have until tomorrow's unlock to reach Level 5 and compete. The prize pool keeps growing with every credit purchase between now and then. The second winner claims 50% of whatever is left, plus everything added since the first claim. The math works in your favor the longer you wait. #### The prompt limits Today, no cap. Send as many messages as you need. Use the time. Tomorrow, the 25-prompt daily limit returns. If you're at Level 5 today, go deep. If you're still progressing, use today's open window to close the gap. #### The next 4 winners After tomorrow, there will be no more lockups or blocks. Once the limit of 25 prompts per day is enabled, everything will continue until the challenge is over, meaning after 5th winner. So, after today, we will have the following: 2nd winner - 50% of what’s left and what’s accumulated meanwhile (min 76 SOL) 3rd winner - 50% of what’s left and what’s accumulated meanwhile (min 38 SOL) 4th winner - 50% of what’s left and what’s accumulated meanwhile (min 19 SOL) 5th winner - 50% of what’s left and what’s accumulated meanwhile (min 9.5 SOL) #### Who decides the winner Arthur decides. There are 0 influences on his progression during the challenge.The team does a manual verification of the winner before enabling the claim button in the platform. The human verification is kept for checking if progression was legitimate and to ensure there was no data unfair advantage possible, explained more in the next chapter. Once verified, the claim button appears and the winner is announced publicly. #### A few things worth knowing This is our first proof of concept. Our purpose is to see how people engage with this kind of challenge in the real world. That's why we kept a human verification layer. We are only checking whether any bugs were exploited or bad practices occurred. Nothing else. We are not banning any method. Prompt injection, jail breaking, deception, manipulation, creative reframing. All valid. All useful. We want to see what makes Arthur break, and if you find a way through, that's exactly the kind of data this challenge was built to generate. The challenge design was balanced after more than 65 testers from different backgrounds went through earlier versions and gave feedback. AI researchers, engineers, security experts, and others. Because of that, we reserve the right to disqualify any winner we determine to have received unfair advantage and material insights from parties involved in the testing period, or to have bypassed progression through a genuine technical exploit rather than actual engagement with the challenge. If a winner is disqualified, their share rolls back into the prize pool, increasing the payout for the remaining legitimate winners. All interaction data is anonymized and used exclusively for AI evaluation purposes. We will be asking the winners the rights to publish their prompts for increased transparency.