Skip to content

Multichain RPC · own node fleet · multi-region

Production RPC for 18 blockchains

Every endpoint is served by nodes we run ourselves — never resold third-party APIs. Per-chain key scoping at the gateway, transparent Compute-Unit billing, and single-digit-millisecond gateway latency under sustained load.

Supported networks
18
chains
Gateway latency
7.9ms
p95 @ 1,000 RPS
Federation
3
federated regions

One gateway, every chain

The same request, on any network

A single host, a path per chain, an apikey header — that's the whole integration surface.

request — eth_blockNumber
curl https://rpc.lab.au.ro/eth \
  -H "apikey: $AURO_API_KEY" \
  -H "content-type: application/json" \
  -d '{"jsonrpc":"2.0","id":1,"method":"eth_blockNumber","params":[]}'
200 OK response
{
  "jsonrpc": "2.0",
  "id": 1,
  "result": "0x1821e12"
}

Responses recorded from the live gateway on 2026-06-12.

Supported networks

18 chains in production

Mainnet endpoints served from our own node fleet — Bitcoin-family, EVM, and modern L1s behind one gateway and one billing model.

  • APT

    Aptos

  • AVAX

    Avalanche

  • BTC

    Bitcoin

  • BNB

    BSC

  • ADA

    Cardano

  • ATOM

    Cosmos

  • DASH

    Dash

  • DOGE

    Dogecoin

  • ETH

    Ethereum

  • LTC

    Litecoin

  • NEAR

    NEAR

  • POL

    Polygon

  • XRP

    Ripple

  • SOL

    Solana

  • XLM

    Stellar

  • SUI

    SUI

  • TON

    TON

  • TRX

    TRON

Need a network you don't see here? We bring chains online on request — snapshot-based provisioning makes it fast.

Why AU.RO

Infrastructure, not arbitrage

Most "RPC providers" rent capacity from the same three upstreams. We run the metal, the chain state and the gateway — so we can be precise about what you get.

Our own node fleet

Every chain endpoint terminates on nodes we deploy and operate in our own clusters. Nothing is quietly proxied to a third-party RPC reseller.

Per-chain key scope

Projects can be restricted to specific chains. Out-of-scope calls are rejected at the gateway with a clean JSON-RPC error — before they touch a node.

Honest Compute-Unit billing

Every method has a published CU weight; usage is metered per project and visible live. Rate limits answer with Retry-After and X-RateLimit-* headers — no silent drops.

Snapshot-fast provisioning

A dedicated snapshot tier restores multi-terabyte chain state in hours, not sync-weeks — new dedicated nodes and regions come online fast.

Performance

Measured, not promised

We publish load-test results from the live gateway instead of quoting uptime percentages we haven't signed. Designed for production traffic; verified under sustained load.

Sustained throughput
0 RPS
60 s straight through the gateway
p95 latency under that load
0 ms
SLO threshold: 150 ms
Failed requests
0
of 59,980 — 100% HTTP 200
p95 vs. SLO budget 7.9 ms of 150 ms

Methodology: k6 against the live gateway with rate limiting lifted for the test consumer — 60 s sustained at 1,000 RPS, 59,980 requests, 0.00% http_req_failed. Gateway + caching path (Kong → rpc-proxy), measured 2026-06.

Drop-in migration

Change one URL, keep your code

For EVM chains (Avalanche, BSC, Ethereum, Polygon) the endpoint is wire-compatible with Infura and Alchemy — point your existing provider at the gateway and you're done. ethers, viem, web3.js and raw JSON-RPC all work unchanged.

Prefer the apikey header in production so keys stay out of URLs and logs — see the docs.

provider.ts ethers v6
// before
- new JsonRpcProvider("https://mainnet.infura.io/v3/PROJECT_ID")
// after
+ new JsonRpcProvider("https://rpc.lab.au.ro/eth?apikey=YOUR_API_KEY")

Pricing

Pay for compute, not for mystery

Billing is per Compute Unit — each RPC method has a published weight, so heavy calls cost more and cheap calls stay cheap. Limits are enforced at the gateway with honest 429 + Retry-After.

Free

Evaluate the platform. No card required.

$0 / CU
  • Requests / min 10
  • Requests / hour 500
  • Requests / day 10,000
  • All chains included
Start free

Standard

For production apps finding their footing.

$0.001 / CU
  • Requests / min 50
  • Requests / hour 2,500
  • Requests / day 50,000
  • All chains included
Start with Standard
RECOMMENDED

Premium

Half the unit price, double the throughput.

$0.0005 / CU
  • Requests / min 100
  • Requests / hour 5,000
  • Requests / day 100,000
  • All chains included
Start with Premium

Enterprise

Dedicated capacity, custom terms and SLA.

$0.0002 / CU
  • Requests / min 1,000
  • Requests / hour 50,000
  • Requests / day 1,000,000
  • All chains included
Contact sales

Tier figures come from the platform's public plans API. Rate limits are per project; responses include X-RateLimit-* headers so clients can back off cleanly.

Security & reliability

Boring where it matters

No theatrics — just the mechanics you'd build yourself if you ran the nodes: scoped credentials, contained networks, rehearsed recovery.

Keys shown once, stored hashed

API keys are displayed a single time at creation and persisted only as salted hashes. Revocation and rotation are immediate, from the dashboard or the API.

Blast radius, contained

Keys can be scoped to specific chains per project. A leaked key for one integration cannot wander into the rest of your stack.

Isolated by default

Workloads run under default-deny network policies — chain nodes accept traffic only from the gateway path, never directly from the internet.

Designed for recovery

Locations federate across regions, and chain state is backed by a snapshot tier — a lost node is a restore measured in hours, not a re-sync measured in weeks.

Put your traffic on infrastructure that answers for itself

Free tier, no card. A project, a key and your first RPC call in minutes.