// Pricing breakdown
How much does a dedicated Solana node cost?
Short answer: self-hosting a production Solana RPC node with Yellowstone gRPC runs $3,500–5,000/mo all-in once you count bandwidth, high availability, and DevOps time. A managed single-tenant node is $2,900/mo flat — and right now your first month is 50% off ($1,450 to start). Shared plans, where you buy a daily request quota instead of a whole server, cost $99–$249/mo. Here's where the money actually goes.
Self-hosted vs managed — line by line
The hardware lease is the headline number, but it's rarely even half of the true monthly cost of running your own node.
| Cost line | Self-hosted | Subglow dedicated |
|---|---|---|
| Bare-metal hardware lease | $1,500–2,500/mo | Included |
| 10 Gbps bandwidth | $80–250/mo (often metered) | Included, unmetered |
| Hot-standby for HA | +~$1,500–2,500/mo | Talk to us for HA pairs |
| Snapshot storage + egress | $50–150/mo | Included |
| Monitoring + alerting | $50–200/mo + setup | Included |
| DevOps / on-call time | 10–30 hrs/mo of eng time | Handled for you |
| Yellowstone gRPC plugin | DIY build per Agave release | Included, free |
| All-in monthly | $3,500–5,000/mo | $2,900/mo |
Self-hosted ranges reflect typical 2026 bare-metal + bandwidth pricing and assume one engineer spending part-time on ops. Your numbers will vary by region and provider.
The hidden costs of self-hosting
The bare-metal box is the cost everyone quotes — but it's the cost that matters least. A single AMD EPYC node with 1 TB of RAM and dual NVMe leases for roughly $1,500–2,500/mo. The real bill shows up around it.
High availability doubles your hardware. One node means one restart storm, one disk failure, or one bad Agave upgrade away from downtime. Serious operators run a hot-standby, which roughly doubles the hardware line. DevOps is the silent killer: snapshot management, ledger maintenance, and plugin ABI breakages on every Agave release eat 10–30 engineer-hours a month. At a loaded eng rate that alone can exceed the hardware.
For the full operational picture — snapshots, TLS, auth, monitoring — see our self-host Yellowstone guide and the managed vs self-hosted breakdown.
What managed pricing looks like
A managed single-tenant node folds all of the above into one flat number. Subglow's AMD EPYC 7543p is $2,900/mo ($2,436/mo effective on yearly), with the Yellowstone gRPC plugin, 10 Gbps unmetered bandwidth, monitoring, and ops all included. There's no setup fee and no credit metering — a traffic spike never turns into a surprise invoice.
For comparison, Helius dedicated nodes start at $2,900/mo, and RPC Fast lists a 1.15 TB node from $2,700/mo in EU/NA, with a 3-month minimum (published prices, checked 2026-10-19). On yearly billing ours works out to $2,436/mo, and you can try month one at 50% off ($1,450).
Do you even need a dedicated node?
Often, no. A flat-rate shared gRPC + RPC plan (Subglow starts at $99/mo) handles the majority of sniping, copy-trading, and analytics workloads. Dedicated hardware earns its cost only when you need single-millisecond latency, isolation for compliance, or sustained volume a shared endpoint would rate-limit. Walk through the decision in the dedicated vs shared checklist, or compare all tiers on the pricing page.
// Managed vs self-hosted
Running it yourself: control against ops
Running your own validator with the Yellowstone plugin gives you total control, and a lot of unglamorous work. Here is the side-by-side before the line-item table above is spent.
- Total control over hardware, region, and config
- Can colocate next to your bot for HFT latency
- $3,500–5,000/mo all-in (hardware + bandwidth + HA)
- 10–30 eng-hrs/mo: snapshots, ABI rebuilds, on-call
- You own TLS, auth, rate-limiting, abuse handling
- Single-tenant AMD EPYC 7543p, your own endpoint
- Yellowstone gRPC plugin included and maintained
- $2,900/mo flat ($2,436/mo yearly)
- Zero ops — we handle upgrades and monitoring
- 10 Gbps unmetered, no credit metering
The ABI treadmill nobody warns you about
The Yellowstone plugin builds against a specific Agave version, and the plugin ABI is tightly coupled to the validator. Every Agave release can require a recompile and a careful restart, and a mismatch makes the plugin refuse to load outright. Multiply that by a hot standby and you have a recurring ops task that lands on whoever is on call. The full operational checklist is in our self-host Yellowstone guide.
When self-hosting genuinely wins
If you are an HFT or market-making shop that needs the validator in the same rack as your execution engine, the colocation latency advantage can justify the ops burden. The same holds if compliance forbids third-party infrastructure. Outside those two cases the maths usually favours managed, and the line items above are where the difference comes from.
Migrating off a self-hosted endpoint
Because Subglow speaks standard Yellowstone gRPC, moving from a self-hosted node is an endpoint and key change — the same @triton-one/yellowstone-grpc client and your existing filters keep working. Pick your client (Agave or Jito) and region on the dedicated node page.
Solana RPC and dedicated node cost — FAQ
How much does a dedicated Solana node cost per month?
A managed single-tenant AMD EPYC 7543p node from Subglow is $2,900/mo flat (currently 50% off the first month — $1,450 to start), or $2,436/mo effective on yearly billing. Self-hosting comparable hardware runs $3,500–5,000/mo all-in once you add bandwidth, a hot standby for HA, and DevOps time. For comparison, Helius dedicated nodes start at $2,900/mo, and RPC Fast lists a 1.15 TB node from $2,700/mo in EU/NA, with a 3-month minimum (published prices, checked 2026-10-19).
How much does Solana RPC cost?
On Subglow, shared Solana RPC with Yellowstone gRPC on the same key costs $99/mo (Sniper) or $249/mo (Pro). New customers pay $79 and $199 for their first 3 months. A managed dedicated node is $2,900/mo, and self-hosting one runs $3,500–5,000/mo all-in. Shared plans come with daily request caps; a dedicated node has none.
What are the cost differences between shared and dedicated Solana RPC nodes?
A dedicated node costs roughly 12–29 times as much as a shared plan: $2,900/mo against $99–$249/mo on Subglow. The shared price buys a daily quota on servers other customers also use (Pro: 500k RPC / 20k sendTx per day); the dedicated price buys the whole server with no per-request caps. Pay for dedicated when sustained volume, isolation or latency variance justifies it; otherwise shared is the cheaper choice.
Why is self-hosting more expensive than a managed node?
The bare-metal lease is only part of the bill. Add 10 Gbps bandwidth ($80–250/mo), a hot-standby node for high availability (roughly doubles hardware cost), snapshot storage, monitoring, and — the big one — DevOps on-call to handle restart storms and plugin ABI breakages on every Agave release. Once you load-in engineering time, a single self-hosted node is usually $3,500–5,000/mo all-in, before you've shipped a line of product.
What hardware does a Solana RPC node need?
For a production RPC node with Yellowstone gRPC you want a recent server CPU (the EPYC 7543p has 32 cores), 1024 GB of RAM, two fast NVMe drives (one for the ledger, one for accounts), and a 10 Gbps uplink. Consumer SSDs die in weeks from write amplification, and under-provisioning RAM causes the validator to fall behind the cluster head.
Is a cheaper shared RPC plan enough instead of a dedicated node?
For most workloads, yes — a flat-rate shared gRPC + RPC plan (Subglow starts at $99/mo) covers sniping, copy-trading, and analytics without the cost of a dedicated node. You only need dedicated hardware when you have strict single-millisecond latency budgets, compliance requirements, or sustained request volume that a shared endpoint rate-limits. The decision checklist is on the dedicated node page.
Are there setup fees or overage charges?
No. Subglow's dedicated node is flat monthly pricing — no setup fee, no credit metering, no per-request overage. You pay in USDC from your dashboard balance. Bandwidth is unmetered at 10 Gbps, so a traffic spike never produces a surprise invoice.
When does self-hosting actually make sense?
Three cases, and outside them managed wins on total cost of ownership. One, you are an HFT or market-making shop that needs the validator in the same rack as your bot for a single-millisecond latency budget. Two, compliance rules require infrastructure you fully control. Three, you already run validators for other reasons and gRPC is a marginal add-on to an existing ops practice.
What's the ongoing ops burden of self-hosting?
Snapshot management, restart storms, ledger growth, TLS and auth in front of the gRPC port, rate-limiting, and — the recurring one — rebuilding the Yellowstone plugin against each Agave release, because the plugin ABI is tightly coupled to the validator version. Budget 10–30 engineer-hours a month plus on-call coverage, and double the plugin work if you run a hot standby.
Is a managed node slower than self-hosting?
Usually the opposite, unless you colocate. A managed node in a well-chosen region with tuned infrastructure typically beats a self-hosted node, because latency is dominated by network distance and infrastructure quality rather than by who owns the box. Self-hosting wins on latency specifically when the validator sits in the same rack as your execution engine.
Can I move from self-hosted to managed without rewriting code?
Yes. Subglow serves the standard Yellowstone gRPC protocol, so migrating from a self-hosted Yellowstone endpoint is an endpoint and API-key change — the same client library and your existing filters keep working unchanged.
Skip the $3,500/mo of hidden ops.
Managed single-tenant AMD EPYC 7543p, Yellowstone gRPC included, from $2,900/mo. First month 50% off — $1,450 to start.