Project 0 (formerly MarginFi):
a battle-tested codebase, rebuilt after its own founder walked away.
We tore apart Project 0, the Solana prime-broker protocol that grew out of MarginFi, across the same lending-adapted scorecard we've used throughout this series. The underlying lending code, marginfi-v2, is genuinely well-tested: a multi-year audit history, 6 audit reports from named firms including OtterSec and Sec3, and an AA (83/100) safety rating from a detailed third-party source as recently as May 2026. But we want to be direct about what happened to get here. In April 2024, hours after Solana liquid-staking protocol SolBlaze publicly accused MarginFi of withholding token emissions owed to depositors, founder Edgar Pavlovsky resigned, citing disagreement with "the way things have been done internally and externally" and, in a since-deleted post, a pledge to block the protocol's own planned token launch. Over the next one to two days, MarginFi lost between roughly $127 million and $260 million in deposits depending on the specific window measured, and TVL fell from an $882 million peak toward roughly half that within 48 hours. What followed was a prolonged, roughly 18-month period of uncertainty before the team relaunched entirely as Project 0 in 2025-2026, with a genuinely different, more ambitious product thesis: unified cross-venue margin letting users borrow against their entire Solana DeFi portfolio, spanning Kamino, Drift, Save, and eventually Jupiter, rather than maintaining separate, siloed collateral on each platform. TVL today sits around $29 million, a 96% drawdown from the 2024 peak, with most of that decline having happened before the rebrand rather than because of it; and landed on a score the marketing page won't show you.
Our take, up front: Project 0 is the current name for what used to be MarginFi, a Solana lending protocol that ranked among the network's top-three lenders in early 2024 before a genuinely serious governance crisis, a prolonged rebuild, and a full rebrand changed what the product is and who trusts it. Real, disclosed, genuinely strong underlying code: marginfi-v2, the codebase Project 0 still runs on, carries a multi-year audit history with 6 audit reports from named firms including OtterSec and Sec3, and a detailed third-party source (DeFi Sentinel) gave it an AA rating and 83/100 safety score as of May 2026, explicitly citing the "battle-tested" nature of the code under its prior brand. No confirmed direct smart-contract exploit found in our research. What we can't set aside: a real, well-documented, serious 2024 crisis. On April 10, 2024, hours after Solana liquid-staking protocol SolBlaze publicly accused MarginFi of withholding token emissions owed to depositors and dumping governance-allocated tokens, founder and CEO Edgar Pavlovsky announced his resignation, stating directly that he didn't "agree with the way things have been done internally or externally." In a separate, since-deleted post made shortly before his resignation, Pavlovsky had pledged to work to block the protocol's own planned token launch, a detail that suggests the resignation was tied directly to an internal fight over whether MarginFi should issue a token at all. The market reaction was immediate and severe: reported outflows over the following one to two days ranged from roughly $127 million in a single day to over $260 million cumulatively, and TVL fell from an $811-882 million peak toward roughly half that within 48 hours. SolBlaze and the MarginFi team reportedly reconciled shortly after. What followed was a real, disclosed, roughly 18-month period of uncertainty, sometimes described online as "airdrop limbo," before the team relaunched under the Project 0 brand in 2025-2026 with a genuinely different, more ambitious product thesis: a unified, cross-venue margin account letting users borrow against their entire Solana DeFi portfolio across Kamino, Drift, Save, and eventually Jupiter, rather than maintaining separate, siloed collateral on each platform individually. Real, disclosed current scale: roughly $29 million in TVL, a 96% drawdown from the 2024 peak, though a detailed source notes most of that decline happened before the rebrand rather than because of it. We weighted all of it below.
Real, disclosed, genuinely long, multi-year audit history for the underlying marginfi-v2 codebase, with 6 audit reports from named firms including OtterSec and Sec3. Real, disclosed structured third-party assessment: an AA rating and 83/100 safety score from DeFi Sentinel as of May 2026, explicitly citing the "battle-tested" code under the prior brand. No confirmed direct smart-contract exploit found in our research.
Pros
- Multi-year audit history; 6 audit reports from named firms (OtterSec, Sec3)
- AA rating, 83/100 safety score from a detailed third-party source (May 2026)
- No confirmed direct smart-contract exploit found in our research
Cons
- The new prime-broker product layer built atop this code is newer and less independently tested
Real, disclosed, dramatic decline: from an $811-882 million peak in 2024, among Solana's top-three lenders, to roughly $29 million today, a 96% drawdown. Real, disclosed precision worth preserving: most of this decline happened before the 2025-2026 rebrand, tied to the 2024 leadership crisis and the prolonged uncertainty that followed, not a failure of the relaunched Project 0 product specifically. Real, disclosed reach claim (self-reported): Project 0's integrations make roughly 70% of Solana lending TVL accessible via its unified margin system, distinct from deposits held directly.
Pros
- Integrations reportedly make a large share of Solana lending TVL accessible via unified margin
Cons
- Direct TVL has fallen 96% from its 2024 peak, to roughly $29 million
- No longer ranks among Solana's largest lenders by direct deposits
Real, disclosed, serious 2024 crisis: SolBlaze publicly accused MarginFi of withholding token emissions owed to depositors and dumping governance-allocated tokens; hours later, founder Edgar Pavlovsky resigned, citing disagreement with "the way things have been done internally and externally," and had separately pledged, in a deleted post, to block the protocol's own planned token launch. Real, disclosed, dramatic capital flight followed (~$127-260M depending on the window measured). Real, disclosed subsequent ~18-month period of uncertainty before a full rebrand. Real, disclosed eventual reconciliation with SolBlaze.
Pros
- The team did eventually reconcile with SolBlaze and ship a coherent, relaunched product
Cons
- A public bad-faith accusation and a founder's resignation occurred within hours of each other in 2024
- The resignation was tied to an internal dispute over the protocol's own token plans
- Roughly 18 months of uncertainty followed before a full rebrand and relaunch
Real, disclosed, genuinely distinctive prime-broker model: a unified margin account lets users post their entire Solana DeFi portfolio, across integrated venues including Kamino, Drift, and Save, as combined collateral, rather than maintaining separate, siloed positions on each platform.
Pros
- Unified margin across multiple integrated venues, a genuinely distinctive model
Cons
- Still early; not all planned integrations (including Jupiter) were live as of our research
Real, disclosed genuinely meaningful UX simplification for active users: one unified margin account instead of separate overcollateralized positions across each platform, if the integrations work as intended.
Pros
- Genuinely meaningful simplification for users active across multiple Solana lending venues
Cons
- Limited disclosed detail on current, live interface quality this early in the relaunch
Real, disclosed historical strength worth noting as context rather than a current metric: Q1 2025 revenue reportedly reached $88.5 million from $1.7 billion in liquidations, predating the fuller decline. Real, disclosed current utilization data (60-80% for key assets) suggesting reasonable spreads for the liquidity that remains.
Pros
- Reasonable utilization (60-80%) on remaining liquidity suggests healthy spreads
Cons
- Historical revenue figures predate the fuller decline and shouldn't be read as current performance
Real, disclosed genuinely ambitious roadmap: planned extension to perpetual futures margining (Drift, Jupiter, Hyperliquid), a "Pay" product, and multi-venue strategy tools, though mostly still forward-looking rather than fully shipped as of our research.
Pros
- Genuinely ambitious, disclosed roadmap spanning perps margining and new product lines
Cons
- Much of the roadmap remains forward-looking rather than fully shipped
Access only through Project 0's official app, and size any deposit against a still-recovering track record.
Given the real, disclosed 96% drawdown from peak and the relatively young prime-broker product layer, treat this less like an established, top-tier lender and more like a newer platform proving itself again, regardless of how well-audited the underlying code is.
Good code doesn't automatically earn back trust that a founder's own exit helped spend.
We want to separate two things that are easy to blur together here. The marginfi-v2 code is genuinely well-tested, independently audited, and by every technical measure we found, sound; nothing in our research suggests the underlying lending logic itself is the problem. The problem is what happened around it: a public bad-faith accusation, a founder's resignation hours later explicitly tied to disagreement over the protocol's own direction, and a token dispute serious enough that he tried to block it on his way out. Whatever the merits on either side of that dispute, the market's response, hundreds of millions of dollars leaving within 48 hours, was a real, measurable verdict on trust, not on code quality. Project 0's new prime-broker thesis is a genuinely interesting answer to a real problem in Solana DeFi, and we don't think the team should be permanently penalized for a founder's decisions years after the fact. But 96% of the deposits that trust once supported still haven't come back, and we think a fair score has to reflect that a good idea, well-executed technically, is still working to earn back something a governance crisis spent.
The scorecard above is deliberately general. Whether Project 0 is right for you depends heavily on which of these you already are.
The active Solana user with positions across Kamino, Drift, and Save who wants genuine cross-venue margin efficiency
This is exactly where Project 0's distinctive prime-broker model delivers real, demonstrated value over siloed collateral.
The user who values a well-audited, multi-year-tested codebase and is comfortable evaluating the team's history separately
The underlying marginfi-v2 code carries a genuinely strong, independently validated security record.
The depositor who sizes exposure to reflect a still-recovering platform rather than a top-tier, established lender
Given the real, disclosed 96% drawdown from peak, this specific calibration genuinely matters here.
Anyone who wants a top-tier lender by scale, or who weighs leadership history as heavily as code audits
Kamino and SparkLend, both reviewed earlier in this series, offer greater scale and a cleaner recent governance record.
The scorecard covers the headline judgment calls. These four tables cover the specifics we didn't want to bury in prose; the April 2024 crisis timelined precisely, the full TVL trajectory, MarginFi vs. Project 0 compared directly, and our fourteenth lending-protocol comparison entry.
April 10-12, 2024, timelined
| When | What happened |
|---|---|
| Apr 10, morning | SolBlaze publicly accuses MarginFi of withholding token emissions owed to depositors and dumping governance-allocated tokens |
| Apr 10, hours later | Founder Edgar Pavlovsky announces his resignation, citing disagreement with "the way things have been done internally or externally" |
| Shortly before resignation | A now-deleted post from Pavlovsky pledges to work to block any future MarginFi token launch |
| Apr 10-12 | Reported outflows range from ~$127M in a single day to over $260M cumulatively across sources; TVL falls from an $811-882M peak toward roughly half that |
| Shortly after | SolBlaze and the MarginFi team reportedly reconcile |
Sources report slightly different specific dollar figures and time windows for the outflows; we're presenting the range rather than picking a single number as definitive.
The fuller TVL trajectory
| Period | Approximate TVL |
|---|---|
| Apr 1, 2024 (peak) | $811-882 million; top-3 Solana lender |
| Apr 12, 2024 (post-crisis) | ~$524-559 million |
| Oct 2025 (pre-Project 0 launch) | Continued decline amid the extended rebrand period |
| 2026 (post-rebrand, current) | ~$29 million |
A detailed source notes most of this decline happened before the rebrand, during the extended period of uncertainty, rather than as a result of the relaunched Project 0 product underperforming.
MarginFi vs. Project 0
| MarginFi (2024) | Project 0 (2026) | |
|---|---|---|
| Core codebase | marginfi-v2 | Same underlying marginfi-v2 (mrgnLendv2) |
| Product thesis | Standard isolated-market lending | Unified, cross-venue prime brokerage |
| Scale | $811-882M peak | ~$29M direct TVL |
| Governance | Founder-led, pre-crisis | New brand, token planned post-rebrand |
Same underlying lending code, a genuinely different product ambition, and a dramatically smaller current scale; all three facts are true at once and worth holding together rather than picking just one.
Lending protocols, side by side (series continues)
| Aave | Kamino | Morpho | SparkLend | Euler | Jupiter Lend | Maple | Venus | Fluid | Project 0 | |
|---|---|---|---|---|---|---|---|---|---|---|
| Most severe disclosed history | $292M bridge exploit | None found | $18M vault loss | None found | $197-240M exploit, recovered | None on contracts | ~$50-54M in 2022 defaults | Known vuln., exploited twice | 3 incidents (~$8.4M), core unaffected | Governance crisis, not a code exploit |
| Distinctive model | Multi-network V3/V4 | Curator-managed markets | Immutable core + vaults | Aave fork, blue-chip only | Modular EVK/EVC vaults | Smart Collateral/Debt, superapp | Institutional credit | Compound fork | Shared layer, lending+DEX | Cross-venue prime brokerage |
Project 0 is the only protocol in this series whose most severe disclosed history is a governance and trust crisis rather than a code-level exploit, a genuinely different kind of risk than most of this series has covered.
We don't just want to hand you our number; we want to show you how it sits next to what other review desks and comparison sites have published. We're including DeFi Sentinel's structured 83/100 safety score directly, since it's a detailed, methodology-driven assessment, while noting its category weights (governance is 20%, versus a larger share in our own methodology) explain much of the gap with our score.
Our score lands meaningfully below the aggregated industry average; most sources weight the strength of the underlying, well-audited codebase heavily and treat the 2024 leadership crisis as resolved history rather than an ongoing factor in current trust and liquidity, which our methodology weighs more heavily.
| Source | Score | Type |
|---|
Scores compiled by our editorial team from publicly available reviews as of August 2026. "Editorial estimate" means the outlet didn't publish a single numeric score, so we converted their published verdict and sentiment into a comparable 100-point figure. Verify current figures directly with each source before citing them elsewhere.
Yes. MarginFi rebranded fully to Project 0 in 2025-2026. The underlying lending code (marginfi-v2) is the same, but the product has grown into a broader prime-broker offering with a genuinely different thesis: unified cross-venue margin rather than standard isolated-market lending.
We found no confirmed direct smart-contract exploit. The protocol's most severe disclosed history is a 2024 governance and trust crisis, not a code-level security failure.
On April 10, 2024, hours after Solana liquid-staking protocol SolBlaze publicly accused MarginFi of withholding token emissions owed to depositors, founder Edgar Pavlovsky resigned, citing disagreement with "the way things have been done internally and externally." He had separately pledged, in a since-deleted post, to work to block the protocol's own planned token launch.
Reported figures vary by source and time window: roughly $127 million in a single day, and over $260 million cumulatively across the following one to two days. TVL fell from an $811-882 million peak to roughly half that within 48 hours.
Roughly $29 million as of our research, a 96% drawdown from MarginFi's 2024 peak. Most of that decline happened before the rebrand, during the extended period of uncertainty that followed the leadership crisis.
A unified margin account that lets users post their entire Solana DeFi portfolio, across integrated venues like Kamino, Drift, and Save, as combined collateral, rather than maintaining separate, overcollateralized positions on each platform individually.
By every technical measure we found, yes. The marginfi-v2 codebase carries a multi-year audit history with 6 reports from named firms, and a detailed third-party source rated it AA (83/100) as recently as May 2026. The governance crisis was a leadership and trust issue, not a code vulnerability.
A token was planned alongside the Project 0 relaunch in 2025-2026, following roughly 18 months of delay and uncertainty after the 2024 leadership crisis, during which the original founder had pledged to try to block any token from launching at all.
More Reviews
Ledn – Crypto Lending & Borrowing Review
Score: 75/100. Highest CeFi score in this series: 10 straight PoR attestations, explicit no-re-lending pledge, never lost customer funds.
Read More
Aave – Crypto Lending & Borrowing Review
Score: 69.5/100. Category leader hit hardest by the Apr 2026 KelpDAO crisis: $124-230M bad debt, its largest liquidity event yet.
Read More
Compound – Crypto Lending & Borrowing Review
Score: 63.25/100. Pioneer whose 2021 self-inflicted $161.7M COMP bug remains its scar, though its caution spared it in Apr 2026.
Read More
CoinRabbit – Crypto Lending & Borrowing Review
Score: 50.5/100. Fast, no-KYC lending with a great product, but no named founder, no PoR, and vague licensing behind it.
Read More