The problem GASX (BlockSpace-Securities-Exchange) solves

GASX - Ethereum Gas Futures Exchange on Sui

AI-native gas futures exchange on Sui, hedged with Thetanuts

The Idea

GASX makes Ethereum congestion tradeable. It tracks one index - the Ethereum Gas Stress Index (EGSI, 0–1000), that measures blockspace stress (base fee, block utilization, mempool pressure, etc). AI forecasts EGSI; users trade futures on it with USDC.

Problem Statement

Every single day, thousands of builders, traders, and protocols get blindsided by gas. You size a trade, you queue a mint, you schedule a DAO payout; and then the mempool erupts, base fee 10x's in six blocks, and your "$4 transaction" clears at "$60 and my dignity." Ethereum's blockspace is the most valuable real estate in Web3, and it is priced live, with zero forward guidance, for everyone who touches it.

It's a structural, economy-wide volatility tax on the entire on-chain economy, and until now, nobody has built the instrument to price it, forecast it, or hedge it.

Market Gap

Every mature market hedges its core input cost. Airlines hedge jet fuel. Farmers hedge wheat. Exporters hedge FX. On-chain, there is no equivalent for gas - despite gas being the single most universal, unavoidable, and violently volatile cost of doing business in crypto.

That gap is enormous, and it is sitting in plain sight:

• DeFi protocols eat unhedged gas risk on every keeper job, liquidation bot, and rebalance.
• NFT projects watch mint margins evaporate the second a drop goes viral and gas spikes with it.
• DAOs budget treasury operations in a currency (gas) with no forward curve.
• Retail traders - literally anyone who has ever smashed "confirm" during a gas spike - are exposed, unhedged, every single time.

If you have ever watched a gas fee eat your alpha: you are the market. GASX exists because that market has never had a product built for it.

Target Users

• DeFi-native power users & bots - keepers, arbitrageurs, liquidators who eat gas cost as a line item and want to hedge it like any other operating expense.
• NFT & dApp teams - need predictable unit economics around mint/interaction windows, not a coin flip on network conditions.
• DAOs & treasuries - want to budget and hedge blockspace exposure the same disciplined way they'd hedge FX or interest rate risk.
• Quant / degen traders - want a clean, liquid, AI-forecasted volatility product that isn't just "another perp on a coin."
• Anyone who has ever rage-quit a transaction because gas 5x'd mid-confirm. (Yes, you. We see you.)

Business Model

GASX monetizes the exact thing it creates: a liquid market for a previously unpriced risk.

• Trading fees - a basis-point take on every winning EGSI-1H contract opened and settled. Classic exchange economics, applied to a brand-new asset class.
• Protocol-owned hedge spread - GASX's own book captures the spread between what it collects in premium and what it pays to hedge on Thetanuts, turning risk management into a revenue center instead of a cost center.
• Premium AI insights - deeper EGSI forecasting, historical analytics, and Kora-powered "gas advisory" for power users, DAOs, and protocols who want to plan treasury ops around a forward curve, not a guess.
• B2B/API licensing - the EGSI feed itself is a product: wallets, dApps, and gas estimators can license the index and forecast to power their own UX, the way market data terminals license price feeds today.
• Expansion flywheel - every new blockspace market (BASE-GAS, ARB-GAS, BLOB-GAS) reuses the same engine, the same hedge rail, and the same user base, compounding distribution instead of resetting it.

TL;DR: first-mover in an entirely uncontested category - gas derivatives, with a credibly neutral settlement layer, a real hedge rail, and a reliable growth path.

Problem, restated simply. Gas congestion risk is hard to quantify, and there is no direct price discovery for future blockspace stress. Companies need predictable gas costs.

Solution, restated simply. Bet on EGSI. If the bet wins, the payout offsets real gas fees.

Challenges we ran into

1. Sui pulled the rug on us mid-build - JSON-RPC just stopped working.
   Every chain read started failing with "JSON-RPC on public fullnodes has been deprecated." Not a bug we introduced, Sui switched it off network-wide. We migrated the entire read path to gRPC, which turned out to be the right call anyway since it's the same transport the frontend's wallet SDK already used.

2. We built an on-chain oracle publisher and then realized we didn't need one.
   Getting the AI's price feed onto Sui meant wrestling a Python SDK through three separate bugs; a config bootstrap that silently required a file that couldn't exist yet, a protocol mismatch, a missing await that let a transaction silently do nothing. We fixed all three.

3. Our first real options trade on Thetanuts kept dying one step from the finish line.
   We got all the way to signing a real RFQ on Base mainnet, and it failed with "option expiry must be after offer deadline" - we were pricing the hedge's expiry off the wrong timestamp. Fixed that, then hit a second, sneakier bug: the transaction builder itself was never awaited, so it silently did nothing instead of erroring. Once both were fixed, the trade actually landed on-chain.

4. The trading UI and the smart contract quietly disagreed about "the price", so every single order failed.
   The frontend priced trades off the AI's forecast; the contract validated them against the live oracle. Two different numbers, both "correct" in isolation, meant every order tripped a slippage guard. We traced it down to one line and made both sides read from the same source of truth.