GetBlock has raised Compute Unit (CU) limits by up to 75% across every Solana Shared Node plan, effective immediately, with no change in pricing. The announcement, published June 23, 2026, applies automatically to existing paid subscribers with no migration, new contract, or key rotation required.
What Changed: The Full Plan Table
GetBlock's Shared Node line spans six tiers. Every tier received a CU ceiling increase:
| Plan | Monthly price | Previous CU cap | New CU cap |
|---|---|---|---|
| Starter | $49 | 50M | 90M |
| Growth | $99 | 105M | 185M |
| Advanced | $199 | 220M | 385M |
| Scale | $349 | 400M | 700M |
| Pro | $499 | 600M | 1,000M |
| Premium | $699 | 900M | 1,600M |
The top Premium tier moves from 900M to 1.6B CUs, a 78% increase. Developers on paid plans see the new limits reflected in their dashboards without any action on their part.
Why RPC Compute Units Matter for Solana Builders
RPC compute units measure the total computational resources a subscriber's API calls consume across a billing period, separate from Solana's per-transaction CU ceiling of 1.4 million CUs, which governs on-chain execution. When a dApp calls getLatestBlockhash, simulateTransaction, or getMultipleAccounts, each request draws from the plan's CU pool. For transaction-heavy workloads (DEX aggregators routing through dozens of accounts per swap, liquidation bots polling collateral health, real-time data products streaming account state), the CU pool is the binding constraint long before the request-per-second limit becomes relevant.
GetBlock's previous caps handled moderate workloads, but teams scaling into high-frequency patterns often found themselves rationing calls late in the billing cycle or upgrading to dedicated nodes earlier than their budget warranted. GetBlock's CEO cited direct customer feedback as the driver: "Builders should spend their time shipping, not rationing requests."
Competitive Context: EMEA Performance Ranking
GetBlock's Solana nodes hold the top latency ranking in EMEA by CompareNodes, which measures round-trip latencies from 27 global locations, with Frankfurt nodes at 6ms average response times and London nodes at 12ms. GetBlock processes more than 50 billion API requests monthly across its 100-plus supported blockchain networks, per the company's own figures.
The CU increase matters most at that latency tier. Developers who chose GetBlock for EMEA performance now have more headroom for the high-frequency call patterns that low-latency nodes are built for, and higher CU ceilings reduce the pressure to upgrade to a dedicated node before the performance case actually justifies it.
Who This Affects and How to Claim It
Existing paid subscribers receive the upgrade immediately with no dashboard action required. Developers on the Free plan access the new limits by upgrading to any paid tier at existing prices. Lapsed subscribers returning during the grace period get the expanded caps at the same plan prices they left.
GetBlock also introduced mid-tier plans (Growth, Scale, Premium) in June to close the gap between starter allocations and enterprise workloads, and launched Limitless Node in May for teams that want dedicated capacity with no per-request ceiling at all ($3,200/month+). The CU increase applies to the shared-node line; Limitless Node operates on a different model.
This update sits alongside a broader wave of Solana RPC infrastructure improvements this month. Triton One's Cloudbreak cut getProgramAccounts latency by 99% through a Postgres-indexed account storage rebuild, and Helius launched a balance-at API giving developers precise historical wallet balance queries at any past slot. The three announcements together reflect ongoing investment in the RPC layer that underpins Solana's developer tooling stack.
Full plan details and the new CU calculator are available at getblock.io/pricing.
