How We Score AI Tools
Every tool on AIToolTier gets the same treatment. Here is exactly how we evaluate, score, and rank them.
The 4-Criteria Scoring System
Each tool is scored on four criteria, each rated 1-10. The Overall score is the average of all four.
Ease of Use (1-10)
How quickly can a new user get value? We test onboarding flow, interface clarity, documentation quality, and learning curve. A tool that requires a PhD to operate scores lower than one you can figure out in 5 minutes.
Output Quality (1-10)
How good is what it produces? For image generators, we look at visual fidelity and prompt adherence. For code tools, correctness and usefulness. For writing tools, coherence and style. We compare outputs against competitors in the same category.
Value (1-10)
Is it worth the money? We weigh the free tier generosity, paid plan pricing relative to competitors, and whether the output quality justifies the cost. A free tool that's decent can outscore an expensive tool that's slightly better.
Features (1-10)
What can it do? We evaluate the breadth and depth of features, integrations, API access, export options, and platform support. A tool with deep functionality in its niche scores higher than a shallow tool that tries to do everything.
The Tier System (S through F)
The Overall score maps automatically to a tier. No manual overrides, no favoritism.
| Tier | Score Range | Meaning |
|---|---|---|
| S | 9.0+ | Best in class. Exceptional across every criterion. |
| A | 8.0 – 8.9 | Excellent. Strong recommendation with minor tradeoffs. |
| B | 7.0 – 7.9 | Good. Solid choice for most users. Some notable weaknesses. |
| C | 6.0 – 6.9 | Average. Works, but better alternatives probably exist. |
| D | 5.0 – 5.9 | Below average. Significant issues. Consider alternatives. |
| F | Below 5.0 | Not recommended. Major problems across multiple criteria. |
Where We Get Our Data
Every review cites its sources. We pull from:
- ✓Official sites — pricing pages, feature lists, documentation, and changelogs.
- ✓User forums — Reddit, GitHub Issues, Discord servers, and product communities for real-world complaints and praise.
- ✓Review platforms — G2, Capterra, Product Hunt, and TrustPilot for aggregated user sentiment.
- ✓Benchmarks — LMSYS Arena, Artificial Analysis, and published benchmark results where applicable.
- ✓Hands-on testing — We use the tools ourselves. No tool gets a review without someone actually using it.
The Known Issues Section
Every review includes a Known Issues section that tracks real problems reported by users. Each issue is sourced (Reddit, GitHub, G2, etc.) and date-stamped so you know when it was reported. This is not a negative hit piece — it is context that helps you make a better decision.
No Paid Placements
Tools cannot pay to get on this site. They cannot pay for higher scores, better placement, or favorable reviews. We may use affiliate links (clearly disclosed) to fund the site, but affiliate partnerships never influence scores or rankings. A tool with an affiliate program gets the same scoring treatment as one without.
How We Stay Current
AI tools change fast. We run an automated daily check across every tool we cover, looking for pricing changes, new features, new known issues, and score-affecting updates. When something changes, we update the review. Every review shows a "Last updated" date so you know how fresh the data is.
Questions about how we review? Learn more about us or browse all tools.