Ethereum RPC Provider Assessment
Compare RPC provider coverage, block visibility, stale-view exposure, and redundancy within a selected region.Illustrative demo data (not measured provider performance)
- Network
- Ethereum mainnet
- Region
- US-East
- Observation window
- July 1, 2026 00:00 UTC → July 31, 2026 23:59 UTC
- Time basis
- Block timestamp
- Data type
- Synthetic mock data
Monitored providers
Monitored regions
Provider labels are anonymized in demo tables to keep the page easy to scan.
Block Coverage
Shows whether each provider delivered a valid observation for the blocks in the selected window.
Provider Coverage
Share of expected blocks with a valid provider observation.
| Provider | Coverage | Missed blocks |
|---|---|---|
| Provider A | 98.9% | 2,455 |
| Provider B | 96.8% | 7,142 |
| Provider C | 94.2% | 12,946 |
Block Visibility Latency
Shows how far each provider lagged the first valid observation of each canonical block in the QuantLoop network.
Provider Visibility Latency
Provider latency relative to the first valid observation of each block.
| Provider | Median | P95 | P99 | P99.9 | Max lag |
|---|---|---|---|---|---|
| Provider A | 82 ms | 210 ms | 390 ms | 820 ms | +738 ms |
| Provider B | 115 ms | 320 ms | 740 ms | 1.65 s | +1.53 s |
| Provider C | 148 ms | 440 ms | 1.28 s | 3.1 s | +2.95 s |
Stale View Exposure
Shows how often each provider remained behind the freshest observed canonical block beyond each latency threshold.
Provider Stale View Exposure
Share of valid block observations that exceeded each threshold.
| Provider | >250ms | >500ms | >1s | >2s |
|---|---|---|---|---|
| Provider A | 7.4% | 3.2% | 0.4% | 0.1% |
| Provider B | 14.8% | 7.6% | 1.2% | 0.3% |
| Provider C | 23.5% | 12.4% | 3.9% | 0.9% |
Head Replacement & Transient Views
During a head replacement, providers may briefly expose a block hash that is later replaced.
Compare how long those transient views remained visible and how often each provider corrected first.
Transient View Recovery
Exposure duration measures how long a transient hash remained visible before the provider observed the canonical replacement. Fastest correction is the share of observed head replacements where the provider corrected first.
| Provider | Median exposure | p95 exposure | Fastest correction |
|---|---|---|---|
| Provider A | 0.4s | 1.2s | 50.0% |
| Provider B | 1.1s | 2.8s | 33.3% |
| Provider C | 1.9s | 4.1s | 16.7% |
Redundancy & Shared Tail Risk
Compare event coverage, effective latency, and P99 tail overlap across provider pairs.
Pair Performance
For each block, pair latency uses the first observation received from either provider. Coverage and latency gains are measured against the better standalone provider.
| Pair | Coverage | p95 | p99 | p99.9 |
|---|---|---|---|---|
| Provider A + Provider B | 99.41%+0.23 pp | 131 ms55 ms faster | 277 ms91 ms faster | 609 ms139 ms faster |
| Provider A + Provider C | 99.49%+0.31 pp | 118 ms68 ms faster | 258 ms110 ms faster | 580 ms168 ms faster |
| Provider B + Provider C | 99.11%+0.08 pp | 153 ms79 ms faster | 338 ms127 ms faster | 760 ms191 ms faster |
Shared Tail Slowdowns
TAIL THRESHOLD
Of the blocks where at least one provider crossed its own historical its own p95 tail threshold, this shows the percentage where both providers did.
Scroll horizontally to compare provider pairs.
What This Helps You Decide
Use the assessment to select providers, choose deployment regions, and design redundancy that reduces shared risk.
Provider selection
Choose providers with strong coverage, fast block visibility, and low stale-view and transient-view exposure.
Regional deployment
See where provider performance changes materially across regions.
Redundancy design
Choose provider pairs that improve coverage and tail latency without slowing down together.