Avalanche node
An Avalanche node, yours alone. The C-Chain without a queue.
An Avalanche node runs the C-Chain for one customer. Your wallet, bot or backend reads blocks and sends transactions through one private endpoint. A backup node runs beside it, and nobody else shares your limits.
- Full
- $139/mo
- Archive
- $289/mo
- Ready
- 24 h
02Full or archive
Avalanche full node or archive node?
Both answer the same C-Chain calls. What changes is how much of the past stays on disk.
Full node
$139/mo
- Starts from
- A state sync, which loads the current state in hours instead of replaying every block
- Query
- eth_call, eth_getLogs, balances and sending transactions on the C-Chain
- Disk
- About 500 GB of active state, per the Avalanche docs, and it grows over the months
- Good for
- Wallets, bots, payment checks and apps that work with the present
Archive node
$289/mo
- Starts from
- The first block. A sync from genesis takes days, not hours
- Keeps
- Every past state of the C-Chain, so old balances and contract values stay readable
- Disk
- Published at about 12.5 TB on one docs page and 13 TB+ on another, growing about 500 GB a month
- Good for
- Explorers, indexers and audits that read old state
Pick in one question
Does your code ever ask what an account held a long time ago? If never, the full node is the one. If it does, you want the archive.
Sources: Avalanche docs, system requirements, checked October 2026; Avalanche docs, bootstrapping a node, checked October 2026
03How it works
How an Avalanche node works: one program, three paths.
Avalanche ships its node as a single program. The C-Chain, the part your app talks to, runs inside it.
AvalancheGo
The node program. The C-Chain, Avalanche's EVM chain, runs inside it.
Three paths
JSON-RPC on /ext/bc/C/rpc, WebSocket on /ext/bc/C/ws and Avalanche-only calls on /ext/bc/C/avax.
Your endpoint
Reached through nodes.herorpc.com. Other nodes find yours on port 9651.
Unlike Ethereum, there is no second program to wire up. One download does the work.
The same node serves three paths: JSON-RPC, WebSocket and the Avalanche-only calls.
The docs list 9650 as the local API port. Your app goes through your private endpoint instead.
04Know the field
Avalanche disk, with the dates attached.
Avalanche has one node program, so the size depends on the mode, not the client. These are the official figures.
Growth, per the Avalanche docs: archive about 500 GB a month, active with snapshots 150 to 200 GB, active with pruning under 10 GB.
AvalancheGo
- Full
- About 500 GB active; 750 GB to 1 TB with snapshots
- Archive
- 12.5 TB (requirements page); 13 TB+ (bootstrapping page)
- Source
- Avalanche docs, read October 2026
AvalancheGo with Firewood
- Full
- Experimental; cannot state sync
- Archive
- Not published
- Source
- Avalanche docs, read October 2026
The two docs pages round the archive differently, so we show both.
Sources: Avalanche docs, system requirements, checked October 2026; Avalanche docs, bootstrapping a node, checked October 2026
05Avalanche, specifically
Six Avalanche facts that change how you plan.
Most guides online are older than the latest C-Chain upgrade. These are from the current docs.
State sync is the default
The C-Chain starts with state sync. It takes hours, where a sync from genesis takes days. Snapshots are stored every 4096 blocks.
No offline pruning after Helicon
Old guides tell you to prune offline to win disk back. Since Helicon the C-Chain ignores those options, so the disk only grows. Plan a fresh state sync when it gets tight.
Disk speed beats disk size
The docs ask for local NVMe that manages at least 3000 IOPS and advise against cloud block storage.
One inbound port
Peers reach a node on TCP 9651. Your app talks to the API port, which stays closed to the outside.
Three paths, one chain
JSON-RPC at /ext/bc/C/rpc, WebSocket at /ext/bc/C/ws and the AVAX-specific calls at /ext/bc/C/avax.
Firewood is still experimental
The new Firewood state scheme is marked experimental and cannot state sync, so a node on it starts differently.
Sources: Avalanche docs, bootstrapping a node, checked October 2026; Avalanche docs, system requirements, checked October 2026; Avalanche docs, C-Chain API, checked October 2026
06Who rents one
Who rents a dedicated Avalanche node.
All four hit the same wall: a shared endpoint that slows down when everyone else is busy.
Wallets and DeFi front ends
Steady eth_call and eth_getLogs reads on the C-Chain, with nobody else drawing from the same pool.
Trading bots
Bots on C-Chain exchanges want quick, predictable answers. A private node keeps their queue empty.
Indexers and explorers
Reading old state for every address needs the archive node. The full node covers everything recent.
Exchanges and payment tools
Watch AVAX deposits on the C-Chain from a node that only your system queries.
07Price check
An Avalanche node for under half of what Allnodes lists.
Allnodes publishes both prices, so full and archive compare like for like. Other dedicated providers start much higher.
Full node
Archive node
Allnodes: the cheapest dedicated plan on its Avalanche pricing page.
Source: Allnodes Avalanche pricing, checked October 2026
RPCFast
From $2,000 a month, plus infrastructure costs
Its dedicated Avalanche cluster page lists this starting price.
RPCFast, dedicated AVAX nodes, checked October 2026
GetBlock
From $1,000 a month
Its dedicated nodes page gives one starting price across chains.
GetBlock dedicated nodes, checked October 2026
You save each month
$181 full, $351 archive
against the Allnodes list price above
08What you get
Pick your Avalanche node and how long to prepay.
Pick the node type and how long to prepay. The price updates as you choose.
Your Avalanche node
Node type
Term
- A dedicated server sized for Avalanche
- AvalancheGo with the C-Chain, one endpoint
- A backup node, included
- A private endpoint on nodes.herorpc.com
- Your server logins
- No setup fee, ready within 24 hours
$139
a month, full node
Billed monthly
Pay in BTC, ETH, USDT, USDC, BNB, SOL, TRX, LTC, DOGE or XMR.
10Questions
Avalanche node questions
Real questions from node operators, answered plainly.
What is an Avalanche node?
It is a computer running AvalancheGo, the Avalanche node program. Yours keeps a copy of the C-Chain, checks every block and answers your app's requests over RPC.
What does it take to run an AVAX node?
The docs ask for local NVMe with at least 3000 IOPS, a steady connection and an open inbound TCP port 9651. Renting skips all of that: the endpoint appears in your dashboard.
Do I need archive state?
Only if your code reads old state. Wallets, bots and payment checks work on a full node. Explorers and indexers that look far back need the archive.
How big is a full Avalanche archive node?
The Avalanche docs put it at about 12.5 TB on one page and 13 TB+ on another, growing around 500 GB a month. An active node is about 500 GB.
Why does the disk keep growing on a pruned node?
Since the Helicon upgrade the C-Chain no longer supports offline pruning, so old guides that suggest it no longer work. The fix is a fresh state sync.
How long does an Avalanche node take to sync?
Hours with state sync, which the C-Chain uses by default. A sync from the first block takes days.
Which RPC path do I call?
Use `/ext/bc/C/rpc` for JSON-RPC, `/ext/bc/C/ws` for WebSocket and `/ext/bc/C/avax` for calls specific to Avalanche.
What does an Avalanche node cost each month?
Full is $139 a month and archive is $289, and neither has a setup fee. Pay for 12 months in one go and the monthly price drops by 25 percent.
When does my Avalanche endpoint arrive?
Inside 24 hours of your payment. The endpoint and your server logins land in the dashboard.
The C-Chain, all to yourself.
Full $139, archive $289, each billed monthly. A backup node comes with it.