Built for Robinhood Chain

One endpoint.
Every provider.
Zero downtime.

Prism RPC is managed RPC infrastructure for Robinhood Chain. Swap in one URL and your traffic is refracted across every major provider - with circuit breaking, automatic failover and live observability behind it.

your appeth_callprismRPCroute · retry · breakAlchemy41 msQuickNode38 msAnkr52 msChainstackcircuit openPublic node96 ms

Pooling capacity from

AlchemyQuickNodeAnkrChainstack+ public Robinhood Chain nodes

Why Prism RPC?

A single RPC URL is a single point of failure. Robinhood Chain settles tokenized real-world assets - the endpoint underneath cannot be the weak link.

Without pooling

One provider, one fate

  • A single provider rate-limits you the moment volume spikes
  • One upstream outage takes your whole app down with it
  • Latency drifts by region and you find out from users
  • Failed reads need retry logic you have to write yourself
  • No shared view of which upstream is actually healthy
With Prism RPC

Every provider, one URL

  • Requests pooled across every major provider on one URL
  • Circuit breakers isolate a failing upstream in milliseconds
  • Latency-aware routing always picks the fastest healthy node
  • Idempotent reads retried automatically with jittered backoff
  • Per-provider health, latency and error rates in real time
How it works

Light in, spectrum out

Four things happen between your call and the answer, and you configure none of them.

01

One beam in

Point your client at a single Prism URL. It speaks the same JSON-RPC that Robinhood Chain nodes do, so nothing else in your stack changes.

02

Refracted out

Each request is scored against every healthy upstream - latency, error rate, remaining quota - and sent to the one that will answer fastest.

03

Failures contained

A provider that starts timing out trips its breaker and leaves the pool. Traffic reroutes before your users notice, and probes let it back in when it recovers.

04

Everything observed

Every hop is measured. Per-provider latency quantiles, breaker state and success rates stream to your dashboard and to Prometheus.

Drop-in

Nothing to install

Prism RPC is a URL, not a package. It speaks standard Ethereum JSON-RPC, so viem, ethers, wagmi, Foundry and every other Arbitrum Orbit tool work unchanged. Change one string and you are pooled.

  • No SDK, no lockfile churn, no version drift
  • Works from browsers, servers, edge runtimes and CI
  • Same request and response shapes as a bare node
  • Leave at any time by pointing the URL back
See viem, ethers and wagmi setups
terminal
curl https://mainnet.prismrpc.co/v1/YOUR_API_KEY \
  -X POST \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","id":1,"method":"eth_blockNumber","params":[]}'

# {"jsonrpc":"2.0","id":1,"result":"0x1f4a2c"}
chain.ts
import { createPublicClient, http, defineChain } from 'viem';

export const robinhoodChain = defineChain({
  id: 42088,
  name: 'Robinhood Chain',
  nativeCurrency: { name: 'Ether', symbol: 'ETH', decimals: 18 },
  rpcUrls: {
    default: { http: ['https://mainnet.prismrpc.co/v1/YOUR_API_KEY'] },
  },
});

const client = createPublicClient({
  chain: robinhoodChain,
  transport: http(),
});

// Prism picks the fastest healthy provider for every call.
const block = await client.getBlockNumber();

Everything a production endpoint owes you

The resilience work you would otherwise write, test and carry yourself - running on the other side of one URL.

Latency-aware routing

Four strategies, including a p50-weighted router that continuously re-ranks upstreams.

Circuit breakers

Per-provider breakers with half-open probing stop one bad node from cascading.

Automatic retries

Idempotent reads retried across providers with exponential backoff and full jitter.

Active health checks

eth_blockNumber probes every few seconds, plus block-height drift detection.

Rate-limit smoothing

Token buckets per provider keep you under every upstream quota at once.

Request deduplication

Identical in-flight reads collapse into one upstream call and fan the answer back out.

Response caching

Deterministic results - receipts, finalized blocks, chain ID - served from a shared cache.

WebSocket subscriptions

eth_subscribe with automatic reconnection and replay of your active subscriptions.

Prometheus metrics

Latency quantiles, breaker state and per-method success rates on a scrape endpoint.

Batch requests

Standard JSON-RPC batches, split across providers and reassembled in order.

Consistency guards

Reads pinned to a minimum block height so you never see a stale node go backwards.

Scoped API keys

Per-key method allowlists, origin locks and independent rate limits.

Routing

Pick how your traffic splits

Set a strategy per API key. Every strategy skips unhealthy upstreams automatically.

default

Latency-based

Ranks healthy upstreams by rolling p50 and sends each request to the current leader. Best for read-heavy apps.

Round robin

Even distribution across the pool. Predictable load, useful when every provider is on the same plan.

Weighted

You set the split - 70% primary, 30% backup - and Prism honours it while skipping unhealthy nodes.

Random

Uniform random selection. The lightest possible routing decision when every millisecond of overhead counts.

<1 ms
Routing overhead (p50)
5
Upstreams per request pool
99.99%
Endpoint availability target
0
Client libraries to install

Ship on Robinhood Chain without RPC anxiety

Create a key, paste the URL, and let the pool absorb whatever the network does next.