The Moyan EI Score
Every other AI directory ranks tools by traffic, funding or how loudly they market themselves. We rate them on one question instead: does using this leave you more capable, or more dependent? That is the EI score — 1 to 10, applied to all 5,234 tools in the directory. Here is exactly how it is calculated, so you can disagree with it precisely.
Step 1 — the category baseline
A tool starts from the score its category tends to deserve. A flashcard app and a bulk-posting tool are not neutral starting points.
| Category | Base | Why |
|---|---|---|
| Education & learning | 9/10 | The whole point is that the human ends up more capable. |
| Research & knowledge | 9/10 | Good tools here show sources you can check yourself. |
| Chat & LLMs | 8/10 | General assistants can teach or can hand over answers; the user sets which. |
| Data & analytics | 8/10 | You still have to read the result and decide what it means. |
| Productivity & automation | 7/10 | Saves time on the mechanical parts, leaves the calls to you. |
| Coding & development | 7/10 | Speeds up typing; understanding the code is still on the developer. |
| Design & UX | 7/10 | Taste and critique stay human. |
| Image generation | 7/10 | Direction and selection remain a human skill. |
| Writing & content | 6/10 | Easy to outsource the thinking, not just the sentences. |
| Marketing & SEO | 6/10 | Often optimised for volume rather than for the reader. |
| Video & animation | 6/10 | Powerful output, limited feedback back to the maker. |
| Voice & music | 6/10 | Same: high output, low practice value. |
| Everything else | 6/10 | Default baseline when a category has no stronger signal. |
Step 2 — credits
At most two signals apply to any one tool, so a single keyword cannot inflate a score.
+2 practice and feedback
Tutoring, coaching, flashcards, quizzes, deliberate practice — the learner stays in the loop.
+2 self-awareness
Journaling, mood, mindfulness, habits — the core of emotional intelligence.
+1 shows its work
Citations, sources, visible reasoning you can verify instead of trusting blindly.
+1 human-to-human
Improves how people communicate — meetings, feedback, interviews, collaboration.
Step 3 — penalties
-2 hands-off publishing
Auto-posting, fully automated, ghostwriting, 'undetectable' — designed so nobody reads the result.
-2 volume for its own sake
Mass outreach, bulk email, scraping. Volume is where quality and trust erode.
-1 identity generation
Voice cloning, face swap, deepfakes — needs consent and care before anything goes public.
Step 4 — clamp and explain
The result is clamped to 3-10: no tool in the directory is worthless, and none is beyond criticism. Every score ships with the sentence that justifies it, printed on the tool page. When our editors write a full profile for a tool, they can override the calculated score, and their reasoning replaces the automatic one.
What the score is not
- It is not a quality rating. A 5/10 EI tool can be the best in its category at its job.
- It is not a user rating. It is our editorial judgment, applied consistently and published openly.
- It is not a security or privacy audit. Check the vendor's own terms before putting sensitive work into any tool.
- It is derived from each tool's stated purpose and description, not from long-term lab testing. Where we have tested, the tool page says so.
Catalog last reviewed 2026-08-25.
Cite or challenge it
The rubric is public on purpose. If you think a score is wrong, the rubric tells you which step to argue with. Quote it, reuse it with attribution to Moyan AI, or write to us and we will re-score the tool.