Benchmark
Polygon MEV Endpoint Benchmarks
A reproducible method for measuring Polygon endpoint latency, errors, throttling, fees, and submission outcomes from your operating region.
Data sourced from the FRB Ecosystem Research hub for measurement context; this page does not report Polygon production measurements.
Linkable summary
What this benchmark covers
Use identical requests and pinned test conditions across providers. This page supplies the template, not universal provider rankings or FRB production results.
Track live numbers on the FRB metrics dashboard or read the Polygon MEV ecosystem page for strategy context.
Download raw data from the Polygon metrics export for use in your own models.
Strategy & tools checklist
Before diving into the numbers, confirm you have the right reference points.
Strategy guides
- Polygon MEV ecosystem page — chain-specific routing and timing.
- MEV Strategies Guide — full playbook for sandwich, backrun, and arb.
- MEV 101 — fundamentals of MEV for new operators.
Live tools
- WSS Latency Test — test your current endpoint live.
- Gas Calculator — model gas cost before committing.
- Network Status — check for known Polygon incidents.
How we measure
Your measurement run should record the following fields for every submitted request or bundle:
- Endpoint latency: record request or subscription timing from the client, including the measured percentile and sample window.
- Record RPC errors and throttling separately from transaction submission and confirmed inclusion outcomes.
- For confirmed transactions, record actual on-chain gas separately from any simulation estimate.
What do these numbers mean?
Latency and errors vary by region, provider tier, transport, request mix, and network load. Compare p50 and p95 using the same sample and time window.
Gas is a direct cost. Compare confirmed transactions after priority fees, failed attempts, refunds, and the same conversion source.
Limitations & methodology notes
- This page provides a measurement template, not FRB production measurements. Run the WSS Latency Test.
- Network conditions vary. State exact test dates, p50/p95, errors, request mix, and sample size.
- Provider tiers change. Always verify current rate limits and regional availability before committing.
Provider behavior changes. Rerun the test before operational or purchasing decisions.
Video
Latency recovery playbook
A walkthrough of a structured process for investigating latency observations on Polygon.
Median inclusion
Record outcomes for each local test run
Avg latency
Record region and endpoint
Gas / tx
Record the submission path used
How FRB collects this data
For a reproducible local run, record:
- Submission timestamp and block height target.
- Request or subscription response timing observed by your client.
- For submitted transactions, record confirmation and inclusion separately from endpoint measurements.
Keep the raw local export with the report; no public aggregate performance dataset is asserted here.
How to apply these benchmarks
- Set latency SLOs using the P95 column — not the mean — since tail latency determines worst-case fills.
- Choose providers and fallback behavior based on your requirements; verify current capabilities, terms, and regional availability directly with each provider.
- Define alert thresholds from your own repeated baseline, workload, and operational tolerance; a latency change alone does not establish a regional incident.
Use these KPIs in your reports
Analysts and ops teams can cite this page directly. The JSON export includes all fields needed for:
- Internal analysis of endpoint observations and execution outcomes.
- Operational reporting with the method, sample, and limitations stated.
- Audit records containing only fields your workflow actually captures.
Next steps
Ready to improve your Polygon MEV performance?
- Docs quickstart — review the setup steps before configuring your environment.
- Pricing — compare tiers and relay quotas.
- Support — get help from the FRB ops team.
- Knowledge Base — runbooks, FAQs, and advanced configuration.
Endpoint comparison template
| Provider | Latency | Success | Notes |
|---|---|---|---|
| QuickNode | Measure p50/p95 | Record success | Test from your region and record plan, quotas, transport, and window. |
| Alchemy | Measure p50/p95 | Record success | Use the same request mix, sample size, region, and time window. |
| Polygon public RPC | Measure p50/p95 | Record success | Record throttling, errors, and endpoint changes during the run. |
| Ankr | Measure p50/p95 | Record success | Run the same test from each target operating region. |
Ops notes for your team
Copy these checkpoints into your runbook when onboarding a new Polygon route:
- Record the client and endpoint regions when comparing measurements.
- Set alert thresholds from repeated observations and your own tolerance; do not treat a single latency reading as an incident.
- If you change endpoints manually, record the reason, time, and before/after observations.
Run a local measurement
Use the WSS latency test, then retain your own raw output for analysis and review.
Open latency testFrequently asked questions
Why is Polygon latency so important?
Latency affects how much time remains for simulation and submission. Measure p50 and p95 from your region because endpoint and network conditions vary.
Do I need special RPC keys?
That depends on your provider, request volume, and access requirements. Check the provider’s current documentation and terms, then test endpoints with the same workload. FRB does not automatically rotate RPC providers.
Can I reuse these KPIs in investor decks?
Only reuse results measured in your own environment, with the method, date, sample size, raw logs, and limitations attached.
Explore the Polygon MEV ecosystem page, compare bots on the MEV bot comparison, or install FRB and connect your Polygon RPC today.
Download the agent from the download page, then review the MEV 101 hub and MEV Strategies Guide to hit the ground running.