99% of protocol revenue automatically goes into the Assistance Fund. It then automatically buys back $HYPE on the open market and permanently burns it.
The official website https://hyperliquid.xyz/ has been updated with the explanation.
Many people still think it’s 97%, but actually it’s 99%.
Real usage = real burn.
This is the core of the $HYPE deflationary flywheel.
Lighter vs Hyperliquid—the difference is very clear.
To become Robinhood’s official perpetual contract partner, Lighter had to pour a large amount of LIT tokens into rewards and incentives—essentially paying upfront to buy the channel, and then also giving points to drive volume.
What about Hyperliquid?
VALR (Africa’s largest crypto exchange) directly integrated Hyperliquid’s permissionless infrastructure using Builder Codes.
No paid tokens, and no heavy commitments.
One is paying to get the entrance.
The other is that, because the technology is open and good enough, it’s actively chosen.
@chameleon_jeff The die-hard again isn’t lurking underwater:
HIP-4 update—simply put, it’s these things
1. You can still add result options after deployment Previously, once you created a prediction market, the results were fixed. Now, deployers can name and add the results afterward.
2. Fees can be adjusted by yourself Deployers can set a fee multiplier (0 to 10x), allowing flexible fee collection.
3. A more user-friendly interface Added shortString hints to make it easier for the frontend to display.
4. Cleaner on-chain status Settlement details no longer occupy L1 state long-term—only those who need them will index them themselves.
5. More testnet templates Old templates will be marked as deprecated.
What’s next: In the next upgrade, prediction markets deployed by validators will officially begin charging fees (about half of ordinary spot).
Early on in mainnet, we’ll be fairly conservative: Each deployer will have at most 100 concurrent markets, with no more than 500 deployments per day. Once stable, we’ll gradually open it up to 1,000 concurrent markets and 5,000 deployments per day.
Overall, this is to make prediction markets more usable and controllable, while managing initial risk.
This is clearly a positive for HIP-4, and also for $HYPE . After fees and the deployer’s share are split, perhaps the remaining 97% will continue to be used for buyback and burn.
Portfolio Margin and Borrow/Lend have reached production-grade scale, have been thoroughly validated, and support meaningful TVL—demand is still growing.
The order book liquidity has also matured; at the current scale, HLP no longer needs liquidity.
Therefore, in the next network upgrade, HLP will begin automatically rebalancing any underutilized USDC into its Lending sub-strategy.
Hyperliquid’s capital efficiency continues to improve.