MutinyFun Docs
The reverse launchpad: retrofit stock-denominated rewards onto coins that already exist. No relaunch, no permission, no lock-up.
What MutinyFun is
A launchpad starts with four blank fields — price, holders, liquidity, mint control — and chooses them. MutinyFun starts with all four already set by someone else. Its job is inheritance: inherit the price, inherit the holders, wrap instead of mint. The customers are the thousands of coins that have a community but no longer a reason to trade — and live coins whose communities want them to pay a dividend.
Anyone can start a mutiny on any qualifying coin. Nobody needs the original team's blessing; a takeover is decided by where holders put their money. Their money, their vote.
Why wrapping, not editing
On Solana a token has no code of its own. The mint is just an account — supply, decimals, two authorities — and all behavior lives in the programs that trade it. Nothing can be added to an existing mint after creation, and on almost every launched coin the authorities are revoked anyway. Even a fully cooperative deployer cannot bolt on new mechanics.
So MutinyFun never touches the original coin. It deploys a permanent two-way wrapper: deposit the coin, receive the wrapped version 1:1, redeem in either direction, forever. Arbitrage welds the two prices together, so a partial migration is not split liquidity — it is one asset with two venues. The wrapped mint trades in a new market against a real stock; the original keeps trading wherever it always did.
How rewards work
- The market: the wrapped coin trades against one deep quote asset. A 1% fee is taken from every trade in that pool — charged in the incoming quote asset, so the mechanism never sells the coin itself.
- Earning is continuous. The wrapper program keeps a running reward index (the Synthetix pattern); every position accrues every second, weighted by the curve below. No snapshots — which also deletes snapshot-timing games.
- Conversion is hourly and permissionless. Anyone can call the harvest-and-convert crank (fees → payout asset via Jupiter, in a Jito bundle) and earns a small caller tip — designed so that no single operator is required for conversions to happen. The platform's keeper is just the politest of the bots.
- Claims are cumulative. Claim any time; one transaction collects everything ever accrued. Unclaimed amounts just sit there.
Every wrapped holder earns — LPing is not required. Holding earns base weight; holding plus a float contribution earns the same stream at boosted weight. Holding and transferring are never taxed in the pool-fee model.
The boarding curve
Early money into an empty vault takes real risk and does real work — it seeds the pool before any fees exist. It is paid for that with a fee-share multiplier priced like a bonding curve:
m(x) = 1 + (M−1)·(1−x)² where x is the fraction of supply already
wrapped and M = 5 is the ceiling (a pilot parameter, not a law of nature).
- Weight = holding + contribution × multiplier. A position's multiplier is the integral of the curve across its entry span — a large entry walks up the curve and gets a worse average, exactly like slippage. Whale-limiting with no cap.
- The state is global, not per wallet — splitting across fifty addresses gains nothing. Sybil resistance for free.
- Forfeit on unwrap: principal is always redeemable 1:1; only the earned multiplier is lost. The contributed float became pool liquidity — there is nothing to hand back.
- The curve rewinds when someone unwraps, so the next entrant buys weight at the true current state.
Fee models & the split
Two models, one rule
- Pool fee (liquid coins): 1% on trades in the new pair only. Holding and sending stay free, so the wrapped coin never trades worse than the original.
- Transfer tax (quiet coins): 1–3% on every transfer, baked into the wrapped mint. For a coin with real volume this drives traders back to the untaxed original — use the pool fee instead. For a coin that barely trades, any capture beats zero.
The split
Starting parameters: of the 1% — 50% to wrapped holders, 25% to the float ratchet (buying pool depth), 25% to the platform, with hard in-program floors of 15% on the holder share and the ratchet. The platform's cut accrues to a platform token via buyback-and-burn, not to any individual — the model established by existing launchpads.
The safety screen
Paste any mint. It qualifies through one of three doors — an established launchpad token, the Jupiter verified list with ≥$50k liquidity, or a pump.fun launch ≥$100k — then passes an automated on-chain audit:
- Freeze authority revoked — non-negotiable. A live freeze authority could freeze the vault itself and break 1:1 redemption. Disqualifier.
- Mint authority revoked — supply fixed.
- No unsafe token extensions, no raisable transfer fee, third-party risk flags.
Governance: parameters, never code
The wrapper program is immutable and identical for every mutiny. The only choices are the ones in the wizard: target coin, paired asset, fee model, and rate. A community takeover is a fight over parameters with no rug surface.
- Curve shape: permanently immutable per instance. It is a promise to everyone who bought weight. Governance here is exit — redemption is always open, so capital can simply leave a configuration it dislikes.
- Fee split: the one adjustable knob, bounded by the in-program floors. Its config authority can be handed to a DAO later without redeploying anything.
- Free two-way redemption is never compromised — no lock-ups, no one-way burns. Continuous exit is what makes permissionless deployment safe.
The MUTINY token
The platform's cut accrues to a token, not a person — the model established by existing launchpads. $MUTINY: fixed supply, authorities revoked. It exists to receive value, and it holds exactly one power.
- Buyback & burn: each epoch the keeper takes the platform's 25% of fees (already in the quote asset), market-buys MUTINY through Jupiter inside a Jito bundle, and burns it — roughly 60% burned, 40% retained for operations. Every burn publishes a receipt on the site.
- Ecosystem flywheel: a slice of the burn budget buys and burns the wrapped tokens of the top mutinies by size — the platform visibly buys its own crews' bags.
- Utility: governance over the fee split — deliberately the only adjustable parameter in the system, still bounded by the hard floors. Nothing immutable is ever votable.
Paired with gold
MUTINY's own pool trades against tokenized gold (GLDx, with Tether Gold as the alternate). The statement is the brand: a pirate platform whose crew is paid in literal gold. Plunder is not a metaphor here — the treasure is treasure. Measured 10 September 2026: GLDx's deepest USDC pool holds $448,310 and every GLDx pool together holds $1,355,551, costing 2.44% on a $250k clip. XAUt0 was not re-measured in that scan, so which of the two is deeper is an open question, not a settled one.
Raydium · Jupiter · Jito
Three integrations, three different superpowers — all public programs and APIs, no forks:
- Raydium — the venue. Concentrated liquidity for the anchored pool (opened at the incumbent price, not discovered from zero), standard fee tiers as the tip jar, and Burn & Earn to permanently lock the float's LP position while still collecting its fees. "The liquidity can never be pulled" becomes a provable on-chain fact, created atomically by the launch transaction.
- Jupiter — the distribution. Once the pool has depth, all aggregated Solana order flow can route through it, and wrapped/original arbitrage runs itself. The swap API also powers the zap: pay in SOL or USDC and it converts on the way in. Verified-list status feeds the safety screen.
- Jito — the execution shield. Scheduled conversions are predictable: a textbook sandwich target. The crank submits swaps as Jito bundles — atomic, front-run-proof. Hourly conversion also means smaller clips: dollar-cost averaging into the payout asset. Idle float can sit in JitoSOL earning staking yield.
Product shape: one on-chain wrapper program, pools on Raydium, one keeper service wearing Jito armor. The keeper is the only operated component.
Measured depth (2026-09-10)
Live Jupiter quotes, USDC into each asset, mints not tickers — price impact by clip size:
| Buying | $10k | $50k | $250k | Verdict |
|---|---|---|---|---|
| SPYx — S&P 500 | 0.02% | 0.05% | 0.36% | Deep |
| TSLAx — Tesla | <0.01% | 0.06% (0.04 to 0.48) | 0.80% (0.66 to 2.58) | Deep, but it wanders |
| GLDx — Gold | 0.12% | 0.62% | 2.44% | Flagged, over 1% |
| Reference memecoin ($ANSEM) | 0.27% | 1.32% (1.24 to 1.37) | 6.26% (6.19 to 6.43) | The real limit |
Quoted on 10 September 2026 between 18:37 and 19:04 UTC. Each figure is the
median of the readings taken in that window; where they spread by a tenth of a point or more,
the band follows in brackets. Mints:
SPYx XsoCS1TfEyfFhfvj8EtZ528L3CaKBDBRqRapnBbDF2W,
TSLAx XsDoVfqeBukxuZHWhdvWHBhgEHjGNst4MLodqsJHzoB,
GLDx Xsv9hRk1z5ystj9MhnA7Lq4vjSsLwzL2nxrwmwtD3re,
$ANSEM 9cRCn9rGT8V2imeM2BaKs13yhMEais3ruM3rPvTGpump.
The finding that matters: the real-asset side trades more easily than the coin does. The stock side was never the bottleneck. Pool-TVL numbers mislead here — quote the route, don't read the TVL.
Gold is the exception this scan found. GLDx costs 2.44% at $250k, past the 1% line the pair list flags on, so it stays on the table marked rather than quietly dropped. Its deepest USDC pool holds $448,310 and every GLDx pool together holds $1,355,551 (DexScreener, by mint, 10 September 2026), which means a $250k clip is a large share of the market it is quoted against. Read the 2.44% as the price of a clip that size, not as a property of the token.
A single quote is not a price. Ten readings minutes apart put GLDx inside a tenth of a point every time, but Tesla came back at 0.66% to 2.58% and SpaceX at 0.51% to 11.61% on the same clip: a route loses its best hop for a few minutes and the cost multiplies. Every figure above is the median of its readings, and any pair whose readings spread by a tenth of a point or more carries the band beside the number rather than instead of it.
Measured and rejected in the same run: AAPLx at 40.62%, METAx at 99.97% with no route at all on five of six tries, and Hylo's leveraged xBTC at 42.68% on a $50k clip with no route on two of six. There is no leveraged bitcoin on Solana that can carry a pool at these sizes. GRND was screened at 80.09% on the 7 September pass and has not been re-measured since.
Two assets on the 7 September scan were not re-measured on 10 September: MSTRx quoted 4.42% at $250k, and the real MSTR share on Backpack filled $10k at 0.08% but returned no fill at $50k. Both figures are three days old and carry that date, not this one.
v1 vs v2
- v1: one payout asset per mutiny — everyone is paid in the pool's quote stock. One swap per epoch, nothing to curate. Transfer-fee extension initialized at rate zero (config authority kept) so leakage protection can be switched on later without a second migration.
- v2: payout elections — each holder picks their stock (S&P, Nvidia, Tesla…) from a capped menu; the keeper aggregates and makes one bulk swap per elected ticker. Purely additive to the distributor; deferred because election menus add real operational surface on the one component that touches money weekly.