All answers
AnswerBacked by Fastest Base WebSocket newHeads push, live block-push lag across RPC providers

Which Base RPC pushes new blocks the fastest over WebSocket?

PublicNode pushes Base newHeads the fastest at 0.4 ms lag versus the earliest provider in the cohort (p50, 24h), measured per block by OpenChainBench across 2 keyless WebSocket providers (PublicNode, dRPC) from a persistent eu-west connection.

Data as of , refreshed continuously.

Base blocks land every 2 seconds, six times more frequently than Ethereum. Any real-time bot on Base, whether a perp keeper, MEV searcher or live dashboard, inherits its provider's head push pipeline six times as often. A multi-second p99 on Ethereum is annoying; on Base it means routinely missing an entire block cycle. This page races the two keyless Base WebSocket endpoints (PublicNode and dRPC) using the same relative scoring as the Ethereum sibling: one persistent eth_subscribe("newHeads") WebSocket per provider from the same eu-west host, earliest arrival per block sets T0, every other provider's lag is arrival minus T0. Tenderly does not expose a keyless Base WebSocket path; mainnet.base.org, Blast, LlamaRPC, BlockPI and Ankr all reject a keyless handshake.

Live leaderboard, top 2

  1. PublicNode logo

    PublicNode

    #1 · Block push lag

    0.4msp99 133 ms
  2. dRPC logo

    dRPC

    #2 · Block push lag

    23msp99 10.24 s

Full live data: /benchmarks/ws-head-latency-base, refreshed every minute.

Methodology and data sources

One persistent eth_subscribe("newHeads") WebSocket per provider from the same eu-west host; each frame is timestamped on receipt. Per block, the earliest arrival defines T0. Each provider's sample is arrival(N) minus T0 in milliseconds. The cohort closes 5 seconds after the first arrival; a provider arriving after that is scored as a missed block. p50, p90 and p99 are computed via Prometheus histogram_quantile over the 24h rate. Base's 2-second cadence generates roughly 43,000 samples per 24h per provider, versus ~7,200 on Ethereum, so the distribution is well-populated. Connection discipline: client ping every 30 seconds, 90-second head watchdog forcing reconnect on silent subscription drop, exponential-backoff reconnects.

What this number does not tell you

  • ·Relative, not absolute. The lag measured is between providers over the same network path, not absolute chain-to-client latency. The ordering and the spread transfer to your integration; the absolute numbers shift with your geography.
  • ·Single vantage point (eu-west). Providers whose infrastructure is US-centric may be disadvantaged relative to what a US-based integration would see.
  • ·Only 2 keyless providers in V1. Tenderly does not expose a keyless Base WebSocket; other gateways reject keyless handshakes. The cohort expands when new keyless paths are verified.
  • ·Win rate and coverage are the robust readings. A provider with a tight p50 that skips blocks loses on the success column rather than hiding in the latency percentiles.

Frequently asked questions

Which Base RPC delivers newHeads first over WebSocket?
PublicNode currently wins most per-block races at 0.4 ms lag versus the earliest provider (p50, 24h) across 2 measured keyless providers. The leaderboard re-scores every ~2 seconds on each Base block.
Why does Base WebSocket latency matter more than Ethereum's?
Base produces a block every 2 seconds; Ethereum every 12. A keeper or MEV searcher wired to a provider that trails by 500 ms misses a meaningful fraction of Base blocks, versus a smaller share on Ethereum's slower cadence. The p99 column is especially important: a multi-second p99 on Base means your app routinely learns about a new block an entire block cycle late.
Why are only 2 providers listed?
V1 measures the keyless tier only: WebSocket endpoints anyone can open without an account. Tenderly does not expose a keyless Base WebSocket path at the time of ship. mainnet.base.org, Blast, LlamaRPC, BlockPI and Ankr all reject a keyless WebSocket handshake. Keyed providers join when contributor keys are wired via environment variables.
Is this the same as Base block time?
No. Base's block interval is fixed at 2 seconds. WebSocket head lag is the additional time between the network producing a block and your RPC provider pushing it to a subscribed client. The lag measured here is relative to the fastest keyless provider observed from eu-west, not absolute.
How does Base WebSocket latency compare to Ethereum?
Cross-chain comparison of absolute lag numbers is not meaningful because block cadences differ. The relative ordering between providers is comparable: if PublicNode leads dRPC by X ms on Ethereum, the gap on Base may differ depending on each provider's Base-specific infrastructure. Use each per-chain leaderboard on its own terms.

Related questions

Same data as /benchmarks/ws-head-latency-base, refreshed every minute. Open methodology, open source.