competitors-analysis
Competitors
1. Jupiter Perps
The biggest DeFi hub on Solana. Has DEX, Perps, Lending/Borrowing, Prediction markets, Wallets, Stablecoin, Stock trading, gamifies rewards, etc. Also has their own token.
Jupiter uses a trader-vs-LP model, where a single liquidity pool acts as the counterparty to every trade. When you win, the pool pays you. When you lose, the pool keeps it. LPs earn 75% of all Jupiter protocol revenue in return for taking on that risk.
Opening a position involves two steps: you submit the order onchain, then an offchain service called a keeper validates and executes it. Pricing is handled by three oracles — Edge (by Chaos Labs) as the primary, with Pyth and Chainlink as backup/validation layers. This multi-oracle setup is a good defense against a single price feed being manipulated.
Supports SOL, ETH, and wBTC, long and short. Only allows six positions in total (one per market). Collateral for longs is the asset itself but shorts require USDC. Order types are limited to market and limit orders, plus TP/SL. Leverage goes up to 250x.
What makes it stand out:
- It's a giant DeFi hub where users can interact with different protocols and mechanisms, one click away.
- The pool earns from trading fees, not token emissions. The APY has historically been strong.
- Recent updates cut the price impact fee by ~60% and doubled max position sizes.
- JLP can be used as collateral in other DeFi products.
Weaknesses:
- Only three underlying assets. No altcoins, no RWAs, no memecoins.
- The keeper model introduces a small latency between order submission and execution.
- The JLP pool is heavily weighted toward SOL.
- You can only hold one position per market at a time, which limits some trading strategies.
2. Flash Trade
Provides Spot and Perps trading, along with several Liquidity Pools and their own token with staking.
Flash Trade runs a pool-to-peer model, but with a key difference from Jupiter: instead of one unified pool, there are multiple pools with different risk profiles, allowing an LP to choose its level of exposure by depositing in a higher or lower risk pool, which is better for liquidity provider experience than "one pool fits all."
Pricing is handled by Pyth, with a risk mechanism on top: if the gap between the current price and the Exponential Moving Average (EMA) exceeds a certain threshold, the platform enters Close Only Mode, where traders can only close positions and remove liquidity. No new opens. This is a circuit-breaker for extreme volatility but it can frustrate traders who want to open in fast-moving markets.
There's also a Degen Mode that pushes leverage up to 500x (from the standard 100x cap).
Flash Trade has two token types: FPL (auto-compounding, tradeable on the market) and sFPL (manual USDC payout every hour). LPs get to pick whether they want passive compounding or active cash flow.
The team is building a second version using MagicBlock's Ephemeral Rollup, an L2 that targets 30-50ms fills.
What makes it stand out:
- Multi-pool system is better for LP risk management than a single pool.
- More asset variety: equities, commodities, memecoins, and DeFi tokens alongside the standard crypto.
- FPL/sFPL split gives LPs flexibility in how they receive yield.
- Degen Mode appeals to a segment of the market.
Weaknesses:
- Until v2 ships, execution speed is comparable to the other LP-model protocols.
- Smaller TVL and volume than Jupiter means thinner liquidity.
- The multi-pool model, while good for LPs, is more complex for traders to understand.
3. Phoenix
Provides perps trading and rewards based on volume traded.
A fully onchain CLOB-based perps exchange. Instead of traders fighting against a pool, it runs a central limit order book (CLOB) where market makers post quotes and traders fill against them. It adds its own innovation with Spline Liquidity, a mechanism that lets market makers define an entire curve of bids and asks using a single mathematical formula instead of posting dozens of individual limit orders. The AMM-style market-making approach drastically reduces the computational cost of updating quotes, which was a major bottleneck for CLOBs previously. Execution is atomic without relying on keepers. It also supports cross margin and isolated margin strategies.
Phoenix doesn't rely on oracles for execution (prices come from the order book itself), but oracles are used as a backstop for margin checks and to trigger TP/SL orders. This removes a point of failure compared with the other perps protocols, where everything is oracle-dependent.
What makes it stand out:
- A real CLOB onchain at competitive costs is something no other Solana perp has pulled off at scale.
- Target fee structure (~5bps) is competitive with Binance, not just other protocols.
- Backed by Paradigm and Haun Ventures, with Anatoly Yakovenko personally involved.
- Trades settle directly onchain, without keepers.
Weaknesses:
- CLOB protocols need active market makers.
- On congested Solana days, high-priority-fee requirements can erode market maker profitability and push spreads wider.
- No token or yield product, so the protocol can't attract passive capital the same way the pool-based protocols can.
4. Adrena
Same trader-vs-LP model as Jupiter, with a single unified liquidity pool. What's different is the philosophy behind the protocol because the team took a deliberate approach to keep the fee model clean and direct, where 100% of revenue goes back to token holders and LPs, no protocol cut, built around a permissionless trading experience and a gamified community layer. Revenue is split 70% to LPs (ALP token holders) and 20% to governance stakers (ADX token holders), with 10% going to a DAO token buyback. Zero to the team.
Adrena has a two-token model: ADX holders vote on the protocol and get a direct cut of fees and ALP holders are the LPs and capture the bulk of trading yield.
Uses oracles Chaos Labs Edge as primary oracle and Pyth as secondary.
What makes it stand out:
- 100% revenue share is a strong incentive and hard to compete with for community building.
- Open source, multiple audits and an active bug bounty on Immunefi.
- Gamified trading experience with regular competitions is a real user acquisition and retention tool.
- Has equities (SPX500, NVDA) and commodities (Gold, Silver, Copper).
Weaknesses:
- Smaller TVL and thinner liquidity, with position sizes capped at $250K per position.
- Fewer markets and still missing major altcoin and memecoin markets.
- Volume dropped suggesting it hasn't yet built a consistent trader base.
- A Q1 2026 audit (Fidesium) found 1 critical and 3 high severity findings (a governance timelock gap was the critical one).
- Team anonymity.
5. GMTrade
Provides the most markets out of any other perps protocol (86+), with an heavy focus on RWAs. Also allows staking of the LP tokens.
GMTrade uses a liquidity pool model, but with two meaningful architectural improvements. First, it uses GM Pools, isolated, per-market liquidity pools so that an LP in a gold market isn't exposed to a memecoin market. Second, on top of those isolated pools sits a Global Liquidity Vault (GLV) that can route idle capital from underutilized pools to where it's needed most, improving overall capital efficiency without collapsing the risk isolation. It has a VOB (Virutal Order Book), that simulates the pools current stake in an order book style, with asks and bids. This is fundammentally a UI/UX feature, not protocol.
For oracles, GMTrade uses Chainlink Data Streams. This is important because Chainlink Data Streams are market-hours-aware, meaning they can actually handle the edge cases that RWA markets require like knowing when markets are closed, when equities halt, etc, preventing stale price exploits during market closures.
On fees,they claim is 5-9x cheaper than other competitors on Solana.
Execution uses a two-step onchain and keeper model. Revenue sharing is 75% to LPs, with 25% retained by the protocol. No token yet but they have a points system.
What makes it stand out:
- The largest asset selection with 86+ markets spanning crypto, forex (EUR/USD, GBP/USD, etc.), commodities (WTI crude, gold, silver, palladium), and equity indices.
- Chainlink Data Streams is one of the most battle-tested oracle solutions, especially for RWA feeds.
- The GLV vault coordination system is a really good capital efficiency mechanism.
- Fees are the lowest of the LP-pool competitors.
- Has hit $554M+ in 24-hour volume at peak, making it consistently one of the top Solana perp DEXs by activity.
Weaknesses:
- Still carries some of the "GMX fork" stigma. Independent credibility is still being earned.
- The team is largely pseudonymous.
6. Pacifica
Provides perps and spot (SOL only) trading protocols with significant markets, several vaults solutions and a small trading game. The most CEX-like protocol of them all. Has some major players involved, like former COO of FTX, OpenAI, Binance, etc.
Pacifica is an offchain order book that separates the matching engine from settlement. The matching engine runs offchain and delivers sub-10ms latency, making it competitive with Binance and Bybit at the execution layer. Settlement stays fully onchain and non-custodial.
The platform supports both cross-margin and isolated-margin, up to 50x leverage, with fees starting at 0.04% for takers. It's also one of the only protocols that ships AI-powered trading tools, described as agents that learn from a user's trading patterns.
The vaults are user created, meaning they open the vaults, define the strategy and manage the vaults afterwards. The rest of the userbase can depotit to provide liquidity and gain passive returns. Very similar to Hyperliquid vaults.
Pacifica's roadmap involves building out a custom Substrate-based L1 to act as the settlement layer, moving toward a full-stack vertically integrated trading infrastructure, similar to Hyperliquid.
What makes it stand out:
- Sub-10ms execution is the fastest of any protocol and competitive with CEXs.
- AI trading tools are a unique feature no competitor has shipped yet.
- User managed vaults, with different trading strategies.
Weaknesses:
- The offchain matching engine is a centralization vector.
- The volume vs. open interest don't add up. At peak, Pacifica had 51% of Solana perp volume but only 14.68% of open interest. The real organic user base is likely much smaller than volume suggests because a lot of that volume is likely airdrop farming
- TVL sits around $27-40M. Small TVL means the protocol can't safely handle large positions without significant slippage.
- FTX association.
Feature Comparison
| Feature | Jupiter | Flash Trade | Phoenix | Adrena | GMTrade | Pacifica |
|---|---|---|---|---|---|---|
| Liquidity model | Single LP pool | Multi LP pool | Fully onchain CLOB | Single LP pool | Multi LP pool + GLV | Offchain CLOB + onchain settlement |
| Max leverage | 250x | 100x (500x w/ Degen Mode) | 25x | 100x | 500x | 50x |
| Markets | SOL, ETH, BTC | Crypto, equities, commodities, memecoins | Crypto, some equities & commodities | Crypto, some equities & commodities | Crypto, forex, commodities, equities, indices | Crypto, forex, commodities, equities, indices |
| Oracle stack | Edge + Pyth + Chainlink | Pyth (EMA circuit breaker ) | CLOB pricing w/ Pyth | Edge + Pyth | Chainlink Data Streams | CEX offchain |
| Execution model | Onchain + Keeper | Onchain + Keeper (V2: rollup) | Onchain | Onchain + Keepers | Onchain + Keeper | Offchain matching + onchain settlement |
| Execution latency | ~400ms | ~400ms (v2: ~30/50ms ) | ~400ms | ~400ms | ~400ms | Sub-10ms |
| Revenue to LPs | 75% | 75% | N/A | 70% | 75% | N/A |
| Taker fee | ~0.06% | Dynamic | ~0.05% | Dynamic | ~0.005% | ~0.04% |
| Gamification | Moderate | Points | None | Strong | Points | Points + Swim game |
| Audit status | Multiple | CertiK, Kudelski | N/A | OtterSec, Fidesium | Zenith | N/A |
| Needs deposit? | No | No (yes for v2) | Yes | No | No | No |
Gaps in the Market
1. Memecoin perps with real liquidity No high liquidity perps for memecoins, which is Solana's biggest retail driver, even though Solana Foundation is trying to divert from it.
2. Cross-chain liquidity aggregation A trader with USDC on Ethereum can't route into these platforms without bridging first, which adds friction and bridge risk. Nobody is natively aggregating cross-chain liquidity into their perps pool. Jupiter is exploring Jupnet (an omnichain aggregation layer).
3. Advanced order types No support for complex trades like OCO (one-cancels-the-other), conditional order chains, etc. Active traders from CEX backgrounds expect these.
4. Copy / social trading Copying trades from successful traders is huge on the bigger CEXs. A protocol that lets users one-click copy a top trader's positions onchain would be a meaningful differentiator, especially paired with gamification.
5. Native mobile-first experience There's no native mobile app for perps, similar to what CEXs have. Solana has the wallet infrastructure to make this real. There's a market segment to be won here.
6. Better LP risk management tools Flash Trade's multi-pool model and GMTrade's GLV system are the best attempts. But LPs still have limited tools as there is no automated delta-neutral vaults built in, no hedging primitives, no transparent real-time LP PnL dashboards.
Points of Failure
Systemic risks across all LP-pool models
If traders have a sustained winning streak, the pool can lose faster than fees can compensate. In a bull run where most perp traders are net long and winning, LPs get hurt badly. None of the protocols have a fully satisfying solution to this beyond closing open interest or raising fees dynamically.
Oracle manipulation
LP-pool models price everything from oracle feeds. A manipulated or stale oracle can let attackers profit at the pool's expense. Jupiter's three-oracle system is the most robust defense. Adrena's dual oracle (Edge + Pyth) is good. GMTrade's Chainlink Data Streams are purpose-built for low-latency DeFi pricing and represent the strongest RWA oracle setup. Flash Trade's EMA circuit breaker is clever but blunt. A sufficiently well-resourced attacker who can move multiple oracle feeds simultaneously is still a theoretical risk across all of them.
Offchain component centralization
Jupiter, Adrena, Flash Trade, and GMTrade all rely on offchain keepers to execute orders. Pacifica takes this further with an offchain matching engine entirely. If keeper infrastructure goes down or gets congested, orders don't execute. If Pacifica's matching servers go down, trading halts completely. Phoenix is the only one that avoids this entirely, but at the cost of CLOB complexity and market maker dependency.
Liquidity fragmentation
The multi-pool models (Flash Trade, GMTrade) solve LP risk management but fragment liquidity across pools. A trader opening a position in a less popular pool might get worse pricing or worse fills than in the main pool. GMTrade's GLV partially addresses this by routing idle capital between pools, but it doesn't fully solve the depth problem for thin markets.
Network congestion
This is a Solana problem. During memecoin mania or major news events, priority fees spike. CLOBs (Phoenix) are hit harder because market makers need to cancel and re-post quotes constantly and those operations consume computation units that get expensive in a fee spike. LP-pool protocols are slightly more resilient because trader order execution is simpler than order book management. Pacifica's offchain matching sidesteps Solana congestion for execution speed, but introduces its own centralization risk in the process.
Opporunities
Oracle design
LP-based perps protocols rely heavily on oracles for price feeds during execution. On the contrary, pure CLOBs like Phoenix avoid oracle dependence for execution entirely, because the market price is discovered solely by matching bids and asks. Oracles in a CLOB are strictly relegated to margin accounting, funding rate calculations, and liquidations.
At first, our CLOB perps protocol wouldnt need oracles, but at a later date we might explore integrating RWAs markets and implementing a framework with an external oracle (e.g., Chainlink) to flag a market halt or extreme deviation, facilitating the management of the order book besides market makers.
Asset selection
For a LP-based perps protocols is easier to spin up new markets, because they only need a new oracle feed and the users trade against the pool. However, for CLOBs it is exponently more difficult due to needing liquidity depth and market makers capital constraints.
We could implement permissionless order book deployments, where we let any team spin up a new market listing with their capital, similar to what Orderly does (see here).
User experience
Pacifica and Phoenix lead on speed (sub-10ms processing), but decentralized order books suffer from UX friction due signing every transaction, managing gas, and deposit/withdrawal latency, etc.
To truly compete, we could implement onchain session keys and ephemeral accounts. A user grants a time limited trading permission to their session, allowing one-click execution without wallet popups, while absolute custody remains in their wallet.
Social trading
Copying the trades of a skilled traders is a big market in CEXs, but its lacking in the DeFi space. The closest thing would the Hyperliquid Vaults, but even then a user is not exactly copying and performing the trade.
In a CLOB protocol, we could implement an atomic copy trade mechanism, where a user would pledge/follow a skilled trader and whenever the trader would submit a order to the order book, the matching engine would create and bundle the follower orders in the same execution block, preventing the followers from getting front-runned by MEV bots.
Gamification & community
Most protocols limit their gamification to their points system and don't really expand on it.
Similarly to what Adrena does, we could organize real competitions between traders, using different metrics (PnL, ROI, etc) to track progress and scores. This would help grow a user-base and keep users engaged. In case of implementing social trading, we could also have "guilds" where the traders form groups and compete between each other.