pyrel
ALL SYSTEMS NOMINALGet an endpoint

MULTI-CHAIN RPC INFRASTRUCTURE

Your node.
Without the node.

Full and archive nodes for 11 chains, run on bare metal we own. One endpoint, every JSON-RPC method, WebSocket subscriptions included. Start in 30 seconds.

Get a free endpoint Read the docs
https://eth-mainnet.pyrel.io/v1/<your-key>
chains live
0
uptime, 90d
0.00%
p50 latency
0ms
edge regions
0

SUPPORTED CHAINS

11 chains. Per-chain truth.

Node type, protocol and method coverage for every network we run. If a namespace is off, the card says so.

01 / 11

Ethereum

MAINNET supportedSEPOLIA supportedHOLESKY supported
FULL supportedARCHIVE supportedTRACE supported
HTTPS supportedWSS supportedgRPC not supported

METHOD FAMILIES

  • eth_call supported
  • eth_getLogs supported
  • eth_getBalance supported
  • debug_traceTransaction supported
  • trace_block supported
  • eth_subscribe supported
  • txpool_content supported
P5038ms

Slide 1 of 11: Ethereum

HOW IT WORKS

Three steps, no infrastructure.

  1. Create a key

    Sign up, pick a chain, get an endpoint URL scoped to one API key. No card, no waitlist.

  2. Point your client at it

    Swap the URL in ethers, viem, web3.py or curl. Same JSON-RPC contract, nothing else changes.

  3. Call any method

    Reads, logs, traces and subscriptions against full and archive nodes. Rate limits are per key, visible in the dashboard.

Everything the node does. None of the pager duty.

Archive depth, trace namespaces and websockets are standard, not add-ons.

ARCHIVE

Archive data from block 0

Historical state at any height on 10 of 11 chains. No separate archive endpoint, no per-call surcharge.

HEAD21,404,881

STREAMS

WebSocket subscriptions, no extra charge

newHeads, logs and pending transactions over WSS on every plan.

DEBUG

Trace & debug namespaces

debug_traceTransaction, trace_block and callTracer where the client supports it.

ROUTING

8-region anycast routing

One hostname resolves to the closest healthy region. Rolling p50, last 60 seconds:

RESILIENCE

Automatic failover across redundant nodes

Every chain runs at least three nodes per region. Unhealthy nodes drain out of rotation in under a second.

VISIBILITY

Usage dashboard + per-method analytics

Requests, errors and latency broken down by method, key and region.

PLAYGROUND

Call it right now.

Pick a method, copy the request in your language. Same shape against every EVM chain we run.

request.sh
curl https://eth-mainnet.pyrel.io/v1/$PYREL_KEY \
  -X POST \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","id":1,"method":"eth_blockNumber","params":[]}'

RESPONSE · 200 OK

{
  "jsonrpc": "2.0",
  "id": 1,
  "result": "0x146f4f1"
}

Eight regions, one hostname.

Requests land in the nearest healthy region. Hover or focus a node for its rolling p50.

LATENCY · eth_call · 24H

Fast at the tail, not just the median.

P50

0ms

P95

0ms

P99

0ms

Pyrel41ms
Provider A64ms
Provider B86ms
Provider C114ms

Illustrative figures — replace with your own measured benchmarks.

Start free. Scale on request volume.

Free

$0forever

Requests
3M / month
Rate limit
25 rps
Archive
1 chain
WSS
2 concurrent
Support
Community
Create a key

Growth

MOST POPULAR

$49/ month

Requests
100M / month
Rate limit
250 rps
Archive
All chains
WSS
50 concurrent
Support
Email, 24h
Start on Growth

Scale

$349/ month

Requests
1B / month
Rate limit
2,000 rps
Archive
All chains + trace
WSS
500 concurrent
Support
Shared Slack
Start on Scale

ENTERPRISE

Dedicated nodes, custom regions, SLA.

Isolated hardware per chain, region pinning, custom rate limits and a signed uptime commitment.

Talk to us
  • Redundant nodes per chain
  • Automatic failover
  • DDoS-filtered edge
  • Status page + webhooks

Questions developers ask.

Point your client somewhere faster.

Get a free endpoint

No credit card. 30-second setup.