Solana’s first network-wide governance vote came down to the final stretch, with SGP-0002, the Double Disinflation proposal, passing by the slimmest of margins.
At the end of epoch 1024, the measure cleared with 67.001% support, only 0.334 percentage points above the 66.667% threshold. Roughly 176 million SOL backed the proposal, about 66.19 million SOL voted against it, and around 20.63 million SOL abstained. A total of 1,326 validators took part, representing 60.7% of eligible stake.
The vote was still losing 70 minutes before the deadline
With 70 minutes left before voting closed, SGP-0002 was still below the passing line, and the no side was ahead by about 58 million SOL.
Kraken 2, a validator linked to Kraken, had switched from yes to no earlier in the final hours, pulling support below the two-thirds threshold. Helius CEO Mert Mumtaz then began a last-minute push. In his own post on X, he said he made about 500 calls in the final few hours.
Then the vote flipped again.
Kraken moved back to support. Galaxy Digital also shifted from abstain to yes. At the same time, JitoSOL holders used Solana’s “staker override” mechanism to bypass validators that had voted no and cast support directly with their own delegated stake weight.
What SGP-0002 changes
The proposal itself is straightforward in design. Solana’s current inflation rate declines by 15% each year. SGP-0002 doubles that disinflation rate to 30%, bringing the 1.5% terminal inflation rate forward by about three years, to around 2029 instead of 2032.
According to the figures in the source article, Solana would issue about 18.9 million fewer SOL over the next six years than previously planned, equal to roughly 2.6% of the current supply. At current prices, that unrealized issuance is valued at about $2 billion.
This is not an overnight supply shock. Inflation would not be cut in half at once; it would decline faster year by year. Estimates from 21Shares show staking yields could fall from about 5.25% to roughly 2.25% over three years. For smaller validators that depend on staking income, that creates direct economic pressure.
Opponents challenged the pace, not the direction
The case against the measure did not reject lower inflation outright. Instead, it focused on the speed of the change and its uneven effect on smaller validators.
Everstake, which cast about 7.96 million SOL against the proposal, said it did not oppose reducing inflation in principle, but questioned how quickly the change would happen and whether the burden would fall disproportionately on small validators. Figment voted about 17.07 million SOL against the measure, making it the largest no voter.
The issue had already failed in earlier forms. In March 2025, SIMD-0228, which pushed in a similar direction, drew about 61% support but fell short of the two-thirds requirement. Small validators mobilized late in that vote and stopped it from passing. SIMD-0411 met the same fate in late 2025 and early 2026.
SGP-0002 was the third attempt. This time, the proposal used a more conservative design, relying on a fixed disinflation rate rather than a dynamic inflation model, and it only just made it through.
Kraken’s 8.9 million SOL became the deciding bloc
The most dramatic part of the vote centered on Kraken 2, a validator tied to the exchange Kraken. The validator controlled about 8.9 million SOL in stake, or around 2% of total voting power.
It first voted yes, then switched to no in the final hours, pushing support below 66.67%. It later flipped back. Protos ran a counterfactual calculation showing that if Kraken’s vote had stayed on the no side, SGP-0002 would have finished with about 63.9% support, well below the passage line.
In effect, a single validator’s change of position altered the network’s token issuance curve for the next six years.
Kraken co-CEO Arjun Sethi responded with a short statement: “custodians should be conduits, not voices.”
Staker sovereignty was used at scale for the first time
Solana’s governance framework under SGP-0001 includes a “staker sovereignty” mechanism. It allows users who delegate stake to override a validator’s vote and express their preference independently with their own stake weight.
That mechanism mattered in the final phase of this vote. Analyst Brian Smith wrote on X that without JitoSOL holders overriding validator positions, SGP-0002 would not have passed either.
The mechanism shows a forward-looking element in Solana’s governance design. But its first large-scale use came in a case where a proposal needed stake holders to reverse validator intent in order to succeed, a detail that will likely remain part of the debate around how voting power should function on the network.
Three proposals were on the ballot
SGP-0001, described as Solana’s constitution, passed easily with 85.97% support. It sets the framework for future governance, including proposal procedures, voting weight and approval thresholds. Under that framework, at least 100,000 SOL in stake is required to submit a proposal.
SGP-0003, the Resource and Inclusion Fee proposal, failed with just 53.9% support, nearly 13 percentage points below the two-thirds threshold.
That proposal would have split Solana’s current fixed signature fee into two parts: a fixed inclusion fee of 2,500 lamports paid to block producers, and a compute-based fee that would be 100% burned. If approved, it could have burned as many as 9,000 SOL per day.
The result matters because it suggests the Solana community is willing to accept “less SOL issuance” on the supply side, but is not yet ready to rewrite fee structure in a major way on the revenue side. Validator incentives diverged across the two measures: lower inflation benefits larger validators that hold more SOL, while fee redistribution would directly affect block-production income.
Helius drafted the underlying technical proposals for both SGP-0002 and SGP-0003, namely SIMD-0550 and SIMD-0553. While pushing for SGP-0003, Mumtaz publicly asked Solana co-founder Anatoly Yakovenko to help rally the remaining validators, but the effort did not produce enough votes.
Passing the vote does not activate the change
For now, SGP-0002 is a governance mandate, not an automatically executed protocol change.
The technical work still needs to move through SIMD-0550 as a separate development process. Client teams must write the code, add a feature gate called double_disinflation_rate, and validators must activate it on mainnet. At the time of publication, SIMD-0550 was still marked “Review” in the Solana Foundation improvement documents repository.
That leaves real distance between authorization and activation. If implementation proceeds smoothly, if the arithmetic logic is accepted across stakeholders, and if the feature gate is activated without dispute, SGP-0002 could stand as a major step toward more active monetary policy management on Solana. If the implementation stalls or the dispute returns, a result that passed by only 0.334 percentage points will be harder to treat as broad consensus.
Solana Company voted no, and the proposal still passed
One detail often missed in the vote is that Solana Company, the entity formerly known as Solana Labs, voted against both SGP-0002 and SGP-0003.
A governance proposal passing even though the network’s core development company opposed it is unusual in blockchain governance. On that point, the outcome sends a positive decentralization signal.
Still, the vote also exposed unresolved issues: who should control the voting power attached to exchange-custodied stake, how much influence large validators should have when they can change position at the last moment, and how far intensive last-minute lobbying by proposal advocates resembles traditional political whip operations.
Solana has now completed one of the crypto sector’s most serious onchain experiments in monetary-policy voting. The result was messy, close and very much alive.