Happy Friday!

Today we look at JTX, Jito's self-custodial trading front end launching July 14, and why building Solana's first institutional-grade workstation matters more than the soft revenue screen suggests. In the Market Update, the week finished broadly green through a whipsaw run of a squeeze, a reversal, and the Hormuz shock, with Modular (+12.0%) leading on Celestia's upgrade and Meme (-15.6%) dragged down almost entirely by MemeCore's M.

The week finished broadly green despite Monday's squeeze, Tuesday's full reversal, Wednesday's Hormuz shock, and Thursday's equity recovery. Modular led all sectors, gaining 12.0%, nearly double the next-best performer. Privacy Index (+6.2%), Lending (+6.1%), Ethereum Ecosystem (+6.1%), and DEXs (+5.5%) rounded out the leaders. Meme was the clear outlier, falling 15.6%, more than four times the decline of the next-worst sector.

BTC gained 2.8%, outperforming the Nasdaq 100 (+1.4%), S&P 500 (+0.8%), and Gold (-0.7%). The weakest crypto sectors reflected idiosyncratic pressures rather than macro beta. Solana Ecosystem (-3.3%) continued to digest last week's liquidation-driven selloff, DePIN (-3.1%) fell after GRASS lowered guidance, and listed crypto equities remained under pressure, with the 2025 Crypto Equity Cohort down 3.5% and Crypto Equities down 1.3%.

Modular's strength came almost entirely from TIA. Celestia deployed its V9 upgrade on July 2, cutting block times from six to three seconds, while its Fibre protocol reached 1 Tb/s of throughput across 498 nodes. TIA rallied 10.1% in a single day as trading volume doubled to $59.5 million.

Meme's 15.6% decline was almost entirely driven by a single token. M fell 24.3% on the week, by far the largest decline in the sector, retracing part of its sharp post-crash rebound. Every other component finished within a much narrower range, from PEPE (+6.7%) to BONK (-8.9%). M's June 25 collapse erased roughly $3 billion in market value on thin liquidity, and the thin float and concentrated supply structure highlighted by ZachXBT in April continues to amplify moves in both directions. Because the index is cap weighted, M's volatility flows directly into sector performance.

— Jake Koch-Gallup

Jito’s JTX Bet

Solana still lacks an institutional-grade trading workstation. Most DEX aggregators remain swap interfaces with adjacent features like DCA and basic limit orders, while professional traders in traditional finance operate from dedicated terminals and execution management systems. The pure front ends that do monetize Solana flow today are built for the opposite client: in June 2026, memes were the majority of front end volume at aggregate net take rates near 60 bps, and less than 1% of trading platform volume came from tokenized assets.

JTX, launching July 14, is Jito's answer: a self-custodial trading front end planning to have market and limit orders, conditional risk orders, and execution tools like TWAPs across spot and eventually perps. The shift is economic as much as product. Jito moves from backend infrastructure, a price taker on blockspace activity, toward owning the interface and becoming a price setter on user flow.

The edge is that Jito can turn Block Assembly Marketplace (BAM) execution primitives into features other front ends cannot easily replicate: larger orders with less information leakage, fresher oracle-dependent pricing, and more reliable conditional order logic. Maker Priority is the first live example, but it is maker-facing, indirectly benefiting users with fresher top-of-batch quotes that should translate into tighter spreads and better fills. The baseline is already competitive, with prior spreads work showing Solana interfaces save users a median 7 to 8 bps round-trip on SOL-USDC versus centralized exchange execution, even against Binance's highest VIP tier.

The prize is bigger than it looks. Dedicated front ends capture only 5 to 10% of Solana DEX volume today, but roughly half of all DEX volume is cyclic arbitrage and market-maker plumbing. Strip that out and genuine user flow approaches half the pie, which is what JTX competes for; it sits at the interface layer while still routing through aggregators, so it can win share without shrinking aggregator volume.

The trade setup exists because Jito has deliberately sacrificed near-term revenue to improve Solana's long-term market structure. Shutting down extractive MEV flows and pushing BAM cut high-margin tip revenue, and softer Solana activity compressed the rest, so reported numbers screen weak. The payoff is a Solana more viable for serious financial activity, especially as tokenized equities and RWAs begin to develop onchain, positioning it for bigger flows as regulatory clarity improves.

That same willingness to trade revenue for a stronger end state is the clearest evidence of management quality. Few crypto teams deliberately give up high-margin income to help their L1 win, and Jito has now made that trade more than once. Now that best-in-class team is building Solana's first credible native trading front end, aimed at the institutional and RWA endgame rather than optimizing for near-term fees, which should help grow the Solana DEX pie instead of just competing for share.

— Sam

The Neobank Trade

The final installment of Blockworks Research's six-part Onchain Banking series adds two adjacent exposures, JUP and AAVE, to the stablecoin-banking basket. Jupiter screens as a trading-first superapp drawing roughly 94% of revenue from perps and its DEX aggregator (about $59M annualized), reinvesting Robinhood-style into cards, lending, and JupUSD, with the Litterbox Trust having deployed about $91.6M into JUP buybacks (281M tokens, ~8.4% of supply). Aave is contested credit-layer exposure: still the lending leader but bleeding deposit share to Morpho after the Kelp incident, with buybacks paused since April 19 and the unreleased Aave App treated as upside rather than base case.

Venice's Burn Thesis

Nikshep Saravanan published a long, self-disclosed-long essay dissecting Venice AI's July 1 raise, arguing the market misread it. Venice closed a $65M Series A led by Dragonfly at a $1B valuation, its first outside capital, while profitable since Q1 on $70M+ annualized revenue and 3.5M users. The core thesis: equity captures the company, but the token captures the surplus, since Venice pays no dividend, runs no equity buyback, and won't sell treasury, leaving VVV buy-and-burn as the only pipe for distributing profit. Saravanan reads the warrants (Dragonfly took part of its consideration in VVV vesting over eight years) as a third-party endorsement of the burn rather than a betrayal. Digging through Venice's production JavaScript, Saravanan found an unannounced "Minds" product (persistent, tool-using custom agents) localized into six languages behind a feature flag, framing a potential agent marketplace as the demand engine that moves both revenue and burn capture at once.