Everything around the endpoint.
Fast RPC is table stakes. What keeps you on a provider is everything around it: keys you can lock down, usage you can see, costs that never surprise you, and a team setup that matches how you work.
Reliability you can hold us to.
Multi-region failover and routing that understands what each call needs. Plumbing you stop thinking about.
Sub-second failover
Requests reroute across regions in under a second, up to five attempts per call. You see a normal response, not an incident.
No gaps, no duplicates
WebSocket subscriptions survive a failover. Missed logs are replayed, reorgs pass through cleanly, and stable subscription ids mean your client never reconnects blind.
Archive and finality aware
Historical queries go to archive nodes, finalized reads to nodes that are caught up, trace calls to nodes that support them. Automatically.
One limit, everywhere
Rate limits are enforced fleet-wide in a single bucket, not per server. What we sell you is what you get, from every region.
Keys that follow your rules.
Scope every key to chains, IP ranges, origins and methods. Give it its own rate limit and daily cap on top of your plan. Rotate with a grace period of an hour, a day or a week, so a redeploy never breaks production.
- IP and CIDR allowlists, enforced at the edge within 30 seconds
- Origin allowlists with wildcards for browser apps
- Allow or deny individual methods or whole families like
debug_* - Per-key rate limits, burst budgets and daily caps
See every request. Keep the receipts.
Break usage down by method, network or key, with human-readable error analytics when something fails. Export request logs as CSV or NDJSON for your own pipeline, and keep 18 months of history without asking us.
Usage & export in the docs →A bill you can compute in advance.
Every method has a published request-unit weight, so you can estimate costs before you ship. Email alerts at 80, 90 and 100 percent of your quota plus one threshold you pick yourself. Overage only happens on-demand, in blocks, under a spending limit you set. There is no bill you did not see coming.
One subscription, your whole team.
Invite teammates by email or link. Decide per member whether they ride on your plan or their own, which chains their keys may touch, how many request units they can spend, and whether they manage their own keys. Governance without a procurement cycle.
Teams in the docs →Built for the way you actually work.
Test any method against any chain in the dashboard playground before you write a line of code. Copy working snippets for ethers, viem, wagmi, web3.py and Go with your key already filled in. Manage keys programmatically over a documented Management API. Even chains without native WebSocket get a synthesized newHeads stream with the same ergonomics.
{ "jsonrpc": "2.0", "id": 1,
"result": "0x147d3b2" } All of it, on every plan.
Nothing on this page is an enterprise add-on. Start a trial and try it against your own traffic.