Solana’s active validator count crossed 2,000 this quarter, up from 1,650 a year ago. The Nakamoto coefficient (the minimum number of validators required to halt the network) is now 32, up from 28. Geographic distribution improved modestly. The critics of Solana’s decentralisation have a thinner argument than they did 12 months ago — but the argument has not gone away.
Key takeaways
- Active validators: 2,047, up from 1,650 in May 2025.
- Nakamoto coefficient: 32 (up from 28), vs Ethereum 4 and Bitcoin 4.
- Top-10 validators control 14% of stake (vs 17% a year ago).
- Hardware requirements remain high — the validator entry-cost barrier is real.
- Active stake concentration in jurisdictions: US 38%, Germany 22%, Switzerland 11%, other 29%.
What the Nakamoto coefficient actually measures
The Nakamoto coefficient is the minimum number of independent validators (or pools) that would need to collude to halt the chain. For Solana, “halt” means stopping block production via >33% offline or malicious stake. The current figure of 32 means 32 validators control 33%+ of stake — anything less and you cannot stop the chain on your own.
For comparison: Ethereum’s Nakamoto coefficient sits at 4 because of staking pool concentration (Lido alone ~28%). Bitcoin’s pool-level Nakamoto coefficient is 4 (Foundry, AntPool, F2Pool, ViaBTC together >50%). On these metrics, Solana is now more decentralised than either at the consensus-participant level.
Where the validators are
Geographic distribution is one of the standard knocks on Solana — historical concentration in a handful of US and German data centres. The 2026 picture is better:
- US: 38% of stake (down from 46% a year ago)
- Germany: 22%
- Switzerland: 11%
- Singapore + Japan: 9%
- Other: 20%
The shift came mostly from APAC validators (Tokyo, Singapore) coming online, plus growth in Switzerland’s data-centre footprint.
Why hardware costs matter
Solana validators run heavy machines — 12+ CPU cores, 256GB+ RAM, multi-terabyte NVMe arrays. Initial setup cost is around $15k–25k and monthly operating cost is in the $400–800 range for the hardware plus data-centre fees. The economic floor for a profitable validator is roughly 30,000 SOL of stake (~$5M at current prices).
This is a real centralisation pressure: only entities with serious capital can run validators. The mitigation is delegated staking (Solana’s native model) — small holders delegate to validators they choose. Whether that constitutes “decentralisation” depends on whether you weight by validator count (good) or by stake-holder count (better).
| Chain | Validator count | Nakamoto coef. | Notes |
|---|---|---|---|
| Solana | 2,047 | 32 | Native delegation |
| Ethereum | 1,067,000+ | 4 | Distorted by Lido concentration |
| Bitcoin (pools) | ~25 active pools | 4 | Top-4 pools >50% of hash |
| Avalanche | 1,700 | ~28 | P-chain validators |
| Sui | 110 | 19 | Smaller validator set by design |
What still concerns the critics
Three persistent concerns:
- Hardware concentration. Most validators run on similar Solana Labs-recommended specifications. A bug specific to that hardware path is a systemic risk.
- Client diversity. The Firedancer client (Jump Trading) is now live but represents a small share of validators. Until Firedancer hits 20%+, the single-client-failure-mode risk remains.
- MEV centralisation. A small number of leader-schedule slots reliably extract the bulk of MEV. The proposed Solana MEV mitigation roadmap is still in design.
“Solana is more decentralised at the consensus layer than its critics give it credit for, and less decentralised at the client and MEV layers than its supporters acknowledge.”
Why this matters for SOL price
Decentralisation perceptions affect institutional acceptance. The same allocators who would not touch Solana in 2023 over centralisation concerns are now putting it in due-diligence pipelines. Whether that translates to flows is a separate question, but the conversation has shifted.
Check the live SOL price + prediction for the model’s current read.
Why this matters
Decentralisation is not a checkbox. It is a multi-dimensional property — consensus, client, MEV, geographic, governance. Solana’s 2026 scorecard is better than its 2024 scorecard. There is meaningful work left on client diversity and MEV. The trend matters more than the snapshot.