How We Test and Rank Web3 Infrastructure

Last updated: October 2026

This page explains how Web3 Infra Guide evaluates infrastructure products and how we decide what to recommend. The same approach applies to every category we cover, from RPC providers to custodians.

What We Look At

Every product is assessed on the criteria that matter for its category. The core criteria are:

  • Fit for the job: whether the product solves the problem the reader has, and for which team size, chain and use case.
  • Reliability and performance: published uptime commitments and service-level agreements (SLAs), status page history, incident reports and, where we can measure it, latency and error rates.
  • Coverage: supported chains and networks, regions, APIs and integrations.
  • Pricing and transparency: whether pricing is public, how it is metered, what free tiers include and how costs scale.
  • Security and compliance: published audits, certifications such as System and Organization Controls 2 (SOC 2) or ISO/IEC 27001 where the vendor lists them, key management model and regulatory status where relevant.
  • Developer and operator experience: documentation quality, SDKs, dashboards, observability and support options.
  • Lock-in and portability: standard interfaces, data export and how hard it is to switch.

Category guides add specific criteria. For example, RPC guides weigh archive data, trace and debug APIs, WebSockets and rate limits, while custody guides weigh segregation of assets, insurance disclosures and licensing.

How We Gather Evidence

  • We read the vendor’s documentation, pricing page, terms and status page, and we cite them.
  • Where a product offers self-serve access, we sign up and test the parts that matter for the guide, and we say what we tested and when.
  • Where a test is quantitative, such as latency, we publish the method: location, date, number of requests and the endpoints used. Small tests from one location are labeled as indicative, not definitive.
  • Where a product is sold only through a sales process, we say that our assessment is based on documentation and public information.
  • Screenshots in our guides are real captures of the vendor’s own pages.

How We Rank

Rankings reflect how well each product fits the use case described in the guide. We explain the reasoning in a verdict under each comparison, including who should choose a different option. Ties and close calls are described as such. Rankings are reviewed when pricing, features or reliability change.

Independence Statement

Rankings and verdicts are decided by our editorial team alone. Vendors cannot pay for a position, a better verdict or the removal of criticism, and they do not see articles before publication. We do not accept tokens, airdrops or allocations connected to coverage. If a guide ever contains sponsored placements or affiliate links, they will be labeled and kept separate from editorial rankings, as set out in our Advertising & Sponsorship Disclosure.

Keeping Guides Current

Each guide shows a “Last updated” date. We recheck pricing, features and key claims when we update a guide, and we note material changes. If you see something out of date, tell us at editorial@web3infraguide.com. Our wider standards are in our Editorial Policy.