APIRank
EN 中文

Editorial Policy

Last updated: July 30, 2026

APIRank is committed to providing accurate, transparent, and useful information about AI API pricing, providers, and services. This Editorial Policy explains how we create, review, and maintain our content.

1. Content Independence

APIRank is an independent information site. We are not affiliated with or owned by any AI API provider. Our rankings, reviews, and recommendations are based on our own analysis and testing criteria, not on commercial relationships.

While we participate in affiliate programs (including the FreeModel referral program), affiliate relationships do not influence our rankings, ratings, or editorial decisions. Providers cannot pay for better placement in our comparison tables or reviews.

2. Pricing Data Verification

Pricing information is the core of APIRank. We take data accuracy seriously:

  • Primary sources: All pricing data is collected from official provider pricing pages, not secondary sources or aggregators.
  • Verification schedule: Major providers (OpenAI, Anthropic, Google, DeepSeek) are verified at least monthly. Other providers are verified quarterly or when we detect a change.
  • Date stamps: Every piece of pricing data carries a "Last verified" date so you know how current the information is.
  • Change tracking: When a provider changes pricing, we log the change with the date it was first noticed and update the comparison tables as soon as possible.
  • Community reporting: Users can report outdated or incorrect pricing via our contact page. Reported errors are reviewed within 48 hours.

Despite our best efforts, pricing can change without notice. We recommend always verifying pricing on the official provider website before making purchase decisions. See our Terms of Use for the full disclaimer.

3. Review Methodology

When we publish in-depth reviews of AI API providers, we evaluate the following dimensions:

  • Pricing competitiveness — How does the provider's pricing compare to alternatives for similar capability levels?
  • Model quality — Evaluation of model output quality, reasoning ability, and consistency across common developer use cases (code generation, content creation, analysis, translation).
  • API reliability — Uptime history, response times, error rates, rate limits, and documentation quality.
  • Developer experience — SDK quality, API compatibility (OpenAI-compatible endpoints), authentication simplicity, and community resources.
  • China accessibility — Whether the service is directly accessible from mainland China, the extent of any restrictions, and available workarounds.
  • Unique value — What differentiates this provider: specialized models, unique capabilities (vision, function calling, streaming), or exceptional pricing on certain tiers.

Not every article evaluates all six dimensions. Quick-reference guides and pricing comparison pages may focus on pricing alone. Full reviews will explicitly state which dimensions were tested and how.

4. Affiliate Disclosure

APIRank uses affiliate links as one of its monetization methods:

  • Some links on the Site are affiliate links. When you click an affiliate link and subsequently sign up for a service or make a purchase, we may receive a commission at no extra cost to you.
  • We disclose affiliate relationships at the bottom of every page (footer) and on the About page. Pages that specifically recommend a service or product include a proximate disclosure near the recommendation.
  • Our current affiliate partners include the FreeModel referral program (invite code FRE-7a3b6220).
  • Affiliate commissions do not influence our rankings, ratings, or editorial decisions. If we recommend a product, it is because we believe it offers genuine value to our readers.

5. AI-Assisted Content Policy

APIRank uses generative AI tools as productivity aids in content creation. Our policy ensures transparency and quality:

  • Human oversight: All content published on APIRank — whether drafted with AI assistance or not — is reviewed, edited, and approved by a human editor (Jason Mai) before publication.
  • Factual verification: All pricing data, provider claims, and technical specifications are verified against primary sources, regardless of how the content was drafted.
  • Transparency: We do not assert that all content is "100% human-written" because our process involves AI tools. However, every piece of content represents our genuine research and editorial judgment.
  • Quality floor: AI-generated or AI-assisted content that is not substantively reviewed and enriched with original analysis is not published. We reject formulaic, templated, or low-effort AI output.

6. Corrections Policy

We strive for accuracy but recognize that errors can occur, particularly in a fast-moving industry like AI API pricing. Our corrections process:

  • Reporting: Errors can be reported via contact@apirank.vip or our contact form.
  • Review timeline: Minor errors (typos, broken links) are typically fixed within 24 hours. Pricing discrepancies are investigated and updated within 48 hours.
  • Significant corrections: If a factual error materially affects the value of an article, we will add a prominent correction notice at the top of the page explaining what was wrong and when it was corrected.
  • Retractions: In the rare case that an article contains fundamental errors that cannot be corrected by revision, the article will be retracted with a clear explanation.

7. Authorship and Attribution

APIRank is primarily operated by a single individual, Jason Mai (about.me/jasonxmai).

  • Articles include author attribution and publication/update dates.
  • When guest contributors or external experts contribute content in the future, they will be clearly credited.
  • No content is published under a pseudonym or aggregated without proper attribution to its source.

8. Conflicts of Interest

We are committed to avoiding conflicts of interest:

  • We do not accept payment from providers for favorable reviews, rankings, or placement in comparison tables.
  • Sponsored content, if ever published, will be clearly labeled as such and will not appear in our comparison tables or ranked lists.
  • We do not accept free API credits, promotional accounts, or other benefits in exchange for editorial coverage. (We may use free tiers or trial credits available to all users for testing purposes, but this does not influence our reviews.)
  • Our affiliate relationships are fully disclosed and operate independently of editorial decisions.

9. Contact

Questions, corrections, or concerns about this Editorial Policy: