Your own ETH/BSC/Polygon RPC endpoint: selling like Alchemy/QuickNode

To create your own ETH, BSC, or Polygon RPC endpoint that can be used as a commercial alternative to Alchemy or QuickNode, you will need a dedicated server with an NVMe drive of at least 2 TB, 64 GB of RAM, and a configured stack consisting of a node (Geth or Erigon) and a reverse proxy with an authorization system—such a configuration, with costs around $200 per month, is capable of serving requests worth up to $3000 per month.

Why run your own RPC endpoint and how to monetize it

The Web3 infrastructure market is overcrowded with centralized giants that dictate high prices and impose strict limits. Setting up your own node allows you not only to save on your own dApps but also to build a full-fledged business. The primary value you sell is low latency, lack of transaction censorship, and high throughput without sudden blocks. When you offer sell rpc services, you target arbitrage bots, MEV searchers, and wallet developers for whom blockchain state update speed is critical.

Project Economics: From Costs to Profit

The business model is built on the difference between hardware rental costs and the cost of API requests. While major providers charge for every "compute unit," you can offer unlimited packages or cheaper credits. An average dedicated server in a Tier III data center costs $150–250. With proper rate-limiting, such a server can handle a load comparable to a competitor's "Growth" plan, which costs between $400 and $900 per network. If you provide access to three networks (Ethereum, BSC, Polygon), your revenue scales proportionally to the number of clients.

Why this is a real alchemy alternative for the market

Many developers are looking for an alchemy alternative due to privacy policies and restrictions on certain JSON-RPC methods. Your own node allows you to provide access to debug_* and trace_* methods, which either cost a fortune or are completely disabled by public providers. This is your main competitive advantage when selling access to specialized teams.

Choosing Hardware for Ethereum RPC VPS and Dedicated Servers

For stable operation, an ethereum rpc vps must have specific characteristics. Regular cloud VPS with slow disks won't handle blockchain synchronization due to high IOPS (Input/Output Operations Per Second). Blockchains constantly write and read data, so using NVMe SSDs is a requirement, not a recommendation.

Characteristic	Minimum (Full Node)	Recommended (Archive/High Load)	Purpose
CPU	4-8 Cores (3.5 GHz+)	16+ Cores (AMD EPYC/Ryzen)	Processing JSON-RPC requests and signing
RAM	32 GB DDR4/DDR5	128 GB+	State caching and database operation
Disk	2 TB NVMe SSD	2 x 4 TB NVMe (RAID 0)	Storing the block chain and indices
Network	100 Mbps Unlimited	1 Gbps+ Port	Peer synchronization and data delivery

Disk Subsystem Specifics

When choosing a server for your own rpc endpoint, pay attention to the TBW (Total Bytes Written) rating of the drives. An Ethereum node can perform petabytes of rewrite cycles in a year. If you plan to run an Archive Node, you will need more than 12 TB of space. For most commercial tasks, a Full Node with pruning enabled (clearing old data) is sufficient, as detailed in the article about Ethereum full node on VPS: Geth + Lighthouse.

Location and Latency

Host servers in key hubs: Frankfurt, Amsterdam, Singapore, or New York. The closer your RPC node is to the main network validators, the faster your clients' transactions will hit the mempool. This is critical for those who want to sell rpc access for high-frequency trading.

Node Installation and Configuration: Geth vs. Erigon

The choice of software determines how efficient your service will be. Geth (Go Ethereum) is the industry standard—it is stable and predictable. However, Erigon (formerly Turbo-Geth) uses disk space much more efficiently and offers higher execution speeds for eth_getLogs queries, making it an excellent quicknode alternative.

Configuring Geth for High Load

To enable the node to serve thousands of external requests, it must be launched with the correct optimization flags. Standard settings are designed for home use, not for a commercial service.

geth --http --http.addr "0.0.0.0" \
--http.port 8545 \
--http.api "eth,net,web3,txpool,debug" \
--http.vhosts "*" \
--http.corsdomain "*" \
--cache 16384 \
--maxpeers 100 \
--db.engine leveldb \
--syncmode snap

The --cache parameter is critical here: allocate at least 25-30% of the server's total RAM to it. This will speed up reading the latest blocks. If you also plan to support other networks, it is useful to study the experience of setting up a Bitcoin full node on VPS, as resource management principles are largely similar.

Erigon: The Choice for a Professional RPC Provider

If your goal is to provide access to historical data (Archive Node), Erigon is unrivaled. It uses a "staged sync" architecture, which compresses blockchain data several times more efficiently than Geth. This allows you to keep an Ethereum archive node on a 3 TB disk instead of 12 TB, significantly lowering the barrier to entry for the business.

Creating Access Infrastructure: Caddy, Rate-limit, and Auth

Simply opening port 8545 to the internet is a surefire way to crash the server instantly. To turn a node into a product, you need a middleware layer that manages access keys and limits the number of requests per second (RPS).

Using Caddy as a Reverse Proxy

Caddy is an ideal choice as it automatically manages SSL certificates and has a modular structure. You can use the rate-limit module to restrict users. The Caddyfile configuration for your own rpc endpoint might look like this:

rpc.yourdomain.com {
    reverse_proxy localhost:8545
    
    handle /v1/api-key-1 {
        rate_limit {
            zone customer1 {
                key {remote_host}
                events_per_second 50
                burst 100
            }
        }
        reverse_proxy localhost:8545
    }
}

Implementing an Authorization System

For commercial use, you need to generate unique paths or tokens for each client. This can be implemented via a custom Authorization: Bearer <token> header or through URL prefixes. If you want to track errors and performance of your middleware in real-time, we recommend deploying Self-hosted Sentry to monitor exceptions in the authorization code.

Monetization and Sales Automation via Stripe

To successfully sell rpc, the payment and key issuance process must be automated. You don't need to write complex billing from scratch. A combination of a simple React/Next.js dashboard and the Stripe API is sufficient.

• Subscription model: Client pays $50/mo for a 10 million request limit.
• Pay-as-you-go model: Using Stripe Metered Billing, where you charge based on actual resources consumed.
• Credit system: The user buys a "credits" package, which is deducted with each API call.

Billing Integration

When a payment is processed through Stripe, your backend should update limits in the database (e.g., PostgreSQL or Redis) and update the proxy server configuration. To analyze user behavior and understand which service packages are most popular, Self-hosted analytics (PostHog or Umami) installed on the same or a neighboring VPS is perfect. This will allow you to see which JSON-RPC methods are used most frequently without violating client privacy.

Marketing Your RPC Service

The easiest way to find your first clients is in developer Discord communities and arbitrage forums. Offer a 24-hour free trial. Since you have your own rpc endpoint, the cost of this trial for you is close to zero, while user trust grows quickly.

Performance Optimization and Monitoring

Clients leave Alchemy not just because of price, but also due to instability during peak network loads (e.g., during major NFT mints). Your task is to ensure a stable uptime of 99.9%.

Server Resource Monitoring

You need to monitor:

1. Disk I/O Wait: If this metric is above 10%, requests will "lag."
2. Peer Count: If the peer count drops to zero, the node will stop syncing.
3. Eth Syncing Status: Checking via eth_syncing that your blockchain head is up to date.
4. Memory Usage: Memory leaks in Geth happen; it's important to set up an automatic restart.

To visualize this data, use the Prometheus + Grafana stack. This is the standard for any ethereum rpc vps. If you plan to scale and launch nodes for other networks (BSC, Polygon), you will face a massive volume of logs. In this case, it's useful to have a structured knowledge base for the team, which can be set up using Self-hosted Outline or BookStack.

Polygon and BSC: Configuration Specifics

Polygon Bor/Heimdall requires much more CPU resources due to high block frequency. BSC (Binance Smart Chain) requires massive disk space and very fast NVMe, as its state size grows faster than Ethereum's. If you plan to offer these networks, budget for servers with at least 128 GB of RAM.

Security of Your RPC Business

Open access to financial infrastructure always attracts hackers. Protection must be multi-layered. Never store private keys for wallets on RPC nodes. The node should be "clean"—only blockchain data.

DDoS Protection and Abuse Prevention

Use Cloudflare in front of your Caddy server to protect against L7 DDoS attacks. Configure your Firewall (UFW or iptables) so that P2P ports (30303 for ETH) are open to everyone, while the node management port (8545) is open only to the local interface or your proxy. If you work in a team, be sure to use Vaultwarden to securely store administrator API keys and server passwords.

Regular Maintenance

Blockchains update frequently (hardforks). You need to subscribe to client developer newsletters (Geth, Erigon, Lighthouse). Missing an update can lead to your node ending up on a fork, and you will be serving incorrect data to clients, which will instantly destroy your service's reputation.

Conclusions

Launching your own RPC service based on Geth or Erigon is a high-margin infrastructure business with a low barrier to entry, where the main asset is a high-quality dedicated server with NVMe. To achieve revenue of $1-3k, it is enough to automate billing via Stripe and ensure low response latency, positioning yourself as an affordable and flexible alternative to major providers.