Review Methodology
This page explains exactly how a tool ends up on Shabelle Hub, how it\u2019s tested, and how the star rating and pricing tier shown on every review are determined. For our broader editorial principles \u2014 sourcing, corrections, and use of AI in our process \u2014 see our Editorial Standards.
1. How We Choose Which Tools to Cover
Real Usage
The tool is widely used, actively discussed, or solves a problem readers regularly search for — not a niche product with no real-world track record.
Actively Maintained
The product is still being developed and supported. We do not list abandoned tools or ones whose pricing pages return errors.
Clear Use Case
We can describe, in plain language, who the tool is for and what task it replaces or improves — if we can’t explain that simply, it doesn’t get a page.
Transparent Pricing
Pricing is published and verifiable on the provider’s own site. “Contact us for pricing” products are noted as such rather than given a misleading price tag.
We currently cover 100 tools under these criteria. Curated coverage means some popular tools may not yet have a page \u2014 not because they failed testing, but because we haven\u2019t tested them yet. See our Contact page to suggest one.
2. The Testing Process
Sign-up & First Impressions
We create an account using the same onboarding flow a new user would experience, including the free tier where one exists. Friction points — confusing setup, required credit cards, unclear plan differences — are noted here.
Core Task Testing
We run the tool through tasks typical for its category: writing and editing for chat assistants, code generation and debugging for coding tools, prompt-to-image generation for image tools, and so on. We pay attention to output quality, speed, and how often results need manual correction.
Limits & Edge Cases
We test what happens at the edges — free-tier usage caps, longer inputs, less common requests — since these are often where marketing claims and real-world experience diverge most.
Paid Tier Comparison
Where a paid plan exists, we compare it directly against the free tier to determine whether the upgrade is worth it, and for whom. This becomes the basis of the “Is it worth paying for?” guidance in each review.
Write-Up & Rating
Findings are written up as pros, cons, use cases, and a numeric rating (see below). The review is checked against the provider’s current pricing and feature pages before publishing.
3. How Comparison Articles Differ from Tool Reviews
Individual tool pages in our directory go through the hands-on testing process described above. Multi-tool comparison articles on our blog synthesize official documentation, published pricing, and feature information for each tool rather than running an identical task across every tool — this is disclosed at the end of each comparison article. Where a comparison references a specific tool's rating, that rating reflects the tool's individual hands-on review.
4. How the Star Rating Is Calculated
Every tool receives a rating out of 5, weighted across five factors observed during testing:
5. What the Score Means
6. What "Free", "Freemium", and "Paid" Mean
Each tool is tagged with one of three pricing labels, used consistently across the directory, category pages, and individual reviews:
Fully usable without payment, with no plan upgrade required for the core feature set.
A genuinely useful free tier exists, with optional paid plans for higher limits, advanced features, or commercial use.
No functional free tier — a paid plan or trial is required to use the product meaningfully.
Updates & Limitations
AI tools change pricing, features, and limits frequently \u2014 sometimes within weeks of a review going live. We periodically re-test listed tools and update ratings and pricing when something has materially changed, but a published date does not guarantee every detail is current at the moment you read it. Always confirm current pricing on the provider\u2019s own site before subscribing.
Ratings reflect our team\u2019s testing and judgment and are inherently subjective \u2014 a tool scored 4.5 here may suit your specific workflow better than one scored 4.8, or vice versa. Use our reviews as one input alongside free trials and your own testing.