What Robinhood Chain's Launchpad Wars Taught Arc

14 September 2026.

Noxa won the opening and did not survive it

Robinhood Chain's mainnet opened on 1 July 2026. Noxa was there, and for eleven days it was effectively the chain's issuance layer — more than 60,000 tokens deployed before it stopped accepting new launches on 11 July, per The Crypto Times' 18 July account. TrustSwap's own Robinhood guide puts the same figure at roughly 60,000 tokens. Its website went dark on 13 July. CoinDesk reported on 15 July that Noxa had generated nearly $12 million in cumulative fees, per DefiLlama, and was redirecting all transaction revenue to creators on the way out.

The part worth reading closely is what happened to the tokens. Noxa's launches were not held in a platform wallet. Each token went into a single-sided Uniswap V3 pool at the 1% fee tier with the liquidity position locked in a locker contract, an arrangement The Crypto Times describes as leaving the team unable to withdraw the pooled funds. When the venue disappeared, the pools did not. CASHCAT, the chain's flagship Noxa launch, kept trading on Uniswap with its liquidity intact — and fell more than 33% in a day and close to three-quarters off its peak in the weeks after, per the same account.

So the tokens were structurally intact and the market priced them as orphans anyway. That is the first finding, and it is uncomfortable for anyone who assumes an on-chain lock is the whole answer. A lock keeps a pool tradable. It does not keep a venue alive, it does not keep a discovery surface pointed at the token, and a permanent lock, on its own, leaves the creator no claim on the LP fee stream inside the position. Whatever a venue chooses to pay creators out of its own revenue — as Noxa did on the way out — lives with the venue and stops when the venue does. What Noxa's holders were left holding was a working pool that nobody had a reason to maintain.

Pons was late and is now the market

Pons was not a day-one venue. It surfaced during Robinhood Chain's first weeks after the 1 July mainnet, and by the end of August it was taking most of the chain's launch activity. The Defiant reported on 1 September that Pons collected $4.89 million in fees on 31 August against $1.72 million for pump.fun, giving it 63.9% of the $7.65 million paid to launchpads across all of crypto that day, with Robinhood Chain recording $1.49 billion in DEX volume over twenty-four hours. A week later The Block reported a record roughly $6 million in daily fees on 3 September, Robinhood Chain weekly DEX volume of $12.4 billion — more than double the prior week — and that roughly 80% of the revenue Pons generates goes to token buybacks, with over 28% of PONS supply burnt.

Two figures, two sources, nine days apart, in the same range. What they support is that Pons cleared several million dollars a day at the end of August and set a record near $6 million on 3 September. What they do not support is a settled daily run rate; both numbers are single-day snapshots taken during an unusually strong week, and anyone quoting "$5–6 million a day" as a standing figure is extrapolating past what was published.

Pons also did not stand still. Its V2, announced on 23 July, replaced the previous launch model with an ETH-denominated bonding curve, graduating tokens at 4.2 ETH into a permanently locked full-range Uniswap V4 position, with creator payouts in ETH by default. The venue that beat the day-one incumbent moved toward a curve, not away from one — which is a fact worth putting next to any argument, including the one below, that curves are the problem.

The lesson is durability, not design

The tempting reading of Robinhood Chain is that Noxa had the wrong mechanism. It did not, particularly. Its locks held, its pools survived it, and it paid out on the way down. What Noxa did not have was a reason to still be there in August, and a launchpad that is not there in August is not a launchpad — it is a set of pools with no maintainer, no front page and no fee stream attaching anyone to their upkeep.

Pons's advantage was not being first and was not, on the evidence, a superior lock. It was being intact and iterating when the chain's activity actually arrived, six to eight weeks after mainnet. Arc's field should read that as a warning about its own composition. Several of the seven venues aiming at Arc were announced within the last month, four of the seven have not published their liquidity treatment, and none has yet processed a mainnet launch; day-eleven market share on a new chain is a measure of who shipped first, not of who will be processing launches in November.

There is a second-order lesson in the CASHCAT outcome that Arc creators in particular should take. Permanent locks and burns both solve the withdrawal problem, and both decide whether the creator's LP fee stream survives the lock — under either, it does not. A venue can still pay creators out of its own revenue, as Noxa did, but that payment lives with the venue, not with the position. That is fine while the venue is healthy and someone else is doing the work of keeping the token visible. It is exactly the wrong structure for the scenario Robinhood Chain actually produced, where the venue leaves and the only party with any residual interest in the pool is the person who launched it.

Where Arc's venues sit

Arc's launch field is split on that question and most of it has not published enough to place. Arch states on its own site that it launches memecoins in single-sided Uniswap V3 pools against USDC with permanently locked liquidity, and does not publish how accrued fees are treated under that lock. Sharc.fun states that liquidity is locked forever at graduation and that creators earn on every trade, without publishing curve parameters, a threshold, or what happens to that fee stream once the liquidity is locked. ARCLaunch, ubi.fun, 5042.fun and AstraPump have published no liquidity treatment this desk can read.

Two of the seven, then, have said something about what the creator keeps: Sharc, in the sentence above, and Bullcheese, which states — per TrustSwap, the operator (Sept 2026) — that the LP fee stream stays claimable by the creator while the liquidity stays locked, and that after 90 days the creator can relock the position and keep the whole fee stream. Bullcheese is a memecoin launchpad built by TrustSwap for Arc, the Layer 1 blockchain created by Circle that uses USDC for gas. It launches tokens single-sided, with no bonding curve and no graduation, and locks liquidity through Team Finance's audited contracts rather than burning it. It is the one venue in the field whose published position pairs a lock with a fee stream that survives it, which is the Noxa scenario's specific gap. It launches on 16 September 2026, the day Arc's public mainnet opens, per TrustSwap, the operator, and no token has launched through it yet. All competitor facts above were verified on each venue's own site on 14 September 2026; row-level status is tracked on Radian's Arc launchpad tracker.

Whether that design survives contact with a real market on Arc is not a question anyone can answer on 14 September, and a venue's published lock terms are not a prediction that the venue will still be operating in December. Robinhood Chain's ten weeks are worth exactly one thing to an Arc creator choosing a venue this week: the question to ask a launchpad is not what it does at launch, but what is left attached to your token when the launchpad is not there.

Sources: The Crypto Times, 18 Jul 2026 · CoinDesk, 15 Jul 2026 · The Defiant, 1 Sep 2026 · The Block, 8 Sep 2026 · crypto.news, 23 Jul 2026 · trustswap.com/robinhood/pons (14 Sep 2026) · thearch.fun, sharc.fun, arclaunch.fun, ubi.fun, 5042.fun, bullcheese.fun (all 14 Sep 2026) · TrustSwap, the operator (Sept 2026).

Bullcheese is a permissionless launch venue. Tokens launched on it are created by anyone, carry no endorsement, and can go to zero. Nothing here is financial advice.

Disclosure: Locksley is built by TrustSwap, and Bullcheese is a TrustSwap product.