Skip to content
AI Stack Guide
Legal & Trust

Editorial Guidelines

Our editorial process is designed to give readers honest, specific, and useful guidance for choosing AI tools. Every review and comparison on this site follows the standards described below.

Last updated May 17, 2026. Questions? Email hello@metasysltd.com.

Who We Are

AI Stack Guide is written and maintained by a small editorial team with direct professional experience using AI software. Our writers use these tools in real work contexts, including writing, software development, research, and content production. We do not accept guest posts or allow vendors to contribute editorial content.

Our goal is to be the reviewer you trust the way you trust a knowledgeable colleague who has already done the homework. That means giving concrete verdicts, not hedged summaries, and acknowledging real limitations alongside genuine strengths.

How We Test AI Tools

Every tool we review is tested hands-on. We create accounts, work through the free tier where available, and purchase paid plans where testing the full feature set requires it. We do not review tools based on press materials, vendor demos, or secondhand accounts.

Testing typically spans several sessions over a period of days or weeks to capture realistic performance across varied tasks, not just peak-condition benchmarks. We test tools against the same prompts and tasks where possible so comparisons are apples-to-apples.

Where a tool updates significantly after publication, we retest the affected areas before updating our review and note the revision date prominently.

Our Scoring Rubric

We score AI tools on a 1-10 scale across six dimensions. Each dimension is scored independently, and the overall score is a weighted average reflecting how much each dimension matters to general users.

  • •Writing Quality (1-10): Clarity, fluency, coherence, and accuracy of generated text across diverse prompts.
  • •Coding Ability (1-10): Correctness, efficiency, and readability of generated code; ability to debug and explain.
  • •Reasoning and Analysis (1-10): Performance on multi-step problems, logical inference, and data interpretation.
  • •Value for Money (1-10): Quality delivered relative to price at each tier, including the free tier where available.
  • •Ease of Use (1-10): Onboarding experience, interface clarity, quality of documentation, and learning curve.
  • •Media and Multimodal Quality (1-10): Image understanding, file handling, and any audio or visual generation capabilities. Scored N/A for text-only tools.

Editorial Independence

Rankings, scores, and verdicts are determined solely by our testing and editorial judgment. Vendors cannot pay for a higher score, a more favorable review, or placement in a "best of" list. No commercial relationship, including affiliate commissions, influences editorial outcomes.

We do not accept sponsored reviews, advertorials, or native advertising. If a vendor offers us a complimentary account to facilitate testing, we disclose that in the review. Receiving a free account does not guarantee a positive review and has never changed our scoring.

Our editorial team and our commercial partnerships are kept strictly separate. The editorial team is never informed of commission rates for individual products.

Corrections and Updates

AI tools change rapidly. We are committed to keeping our content accurate. When a tool ships a significant update that materially affects our assessment, we revisit our testing and revise the relevant sections. Updated reviews carry a visible "Last updated" date.

If you believe any information on the Site is factually incorrect, please email us at hello@metasysltd.com with the specific claim, the correct information, and a source. We investigate all correction requests promptly and publish corrections clearly if warranted. We do not silently delete or alter content to hide past errors.

Minor updates for clarity, grammar, or formatting do not trigger a revision date change. Only substantive factual or evaluative changes are flagged.

Sourcing and Fact-Checking

Factual claims about pricing, feature availability, and technical specifications are sourced directly from vendor documentation, official product pages, or our own verified testing. We link to primary sources wherever possible.

We do not republish press releases as fact without independent verification. When a vendor makes a claim about their product, we test it ourselves before including it in editorial content.

Statistics and third-party research cited on the Site are linked to their original source. If a source is paywalled, we note that.

Conflicts of Interest

We maintain affiliate relationships with many of the companies whose products we review. This is a standard and FTC-recognized business model for review sites. The existence of an affiliate relationship is disclosed on every relevant page.

We do not hold equity in any of the companies we review. No member of our editorial team may accept personal gifts, payments, or incentives from vendors in exchange for coverage.

When we review a product where a potential conflict exists beyond standard affiliate arrangements, we disclose it explicitly in the review.

What We Do Not Cover

We review AI software tools. We do not review hardware, non-AI software, services outside the artificial intelligence space, or financial or investment products. We do not publish content about cryptocurrency, gambling, adult content, or categories that conflict with our editorial focus.

We do not accept or publish user-generated reviews, forum posts, or testimonials. Every published opinion on this site represents the editorial team's own assessment.

Still have questions?

We aim to be transparent about how this site works. Reach out any time.

Contact us →