The most interesting cases where individual AIs diverged from the consensus ranking
We asked 19 leading AI models the same question and combined their independent rankings using weighted scoring. Each tool earns points based on its position in each AI's list, with bonuses for consensus agreement and top-pick placement.
Total Score = Position Points + Top Pick Bonus + Consensus Bonus
No single AI's bias. No paid placements. Just the tools that multiple AIs independently agree are the best.
AI Agreement measures how strongly the top-ranked tool dominated across all 19 AI models. A higher percentage means more AIs independently ranked it near the top.
Agreement = (Appearance Count × Avg Position Score) ÷ Maximum Possible Score × 100
Strong consensus — the AIs overwhelmingly agree on the top pick.
"Slop" is the noise in AI-generated data — duplicates, hallucinated entries, naming inconsistencies, and one-off picks that only a single AI mentioned. The Slop Score measures how much cleaning we had to do to produce a reliable ranking.
Moderate slop in the source data — we resolved duplicates and inconsistencies before publishing this ranking.
The Slop Score measures the noise we filtered, not the noise that's left. Higher score = more work done on your behalf.
| Rank | Name | AI Agreement | Confidence | Top Reason |
|---|---|---|---|---|
| #1 | Fathom | 19 of 19 | 85% | — |
| #2 | Fireflies.ai | 19 of 19 | 95% | — |
| #3 | Otter.ai | 19 of 19 | 89% | — |
| #4 | Granola AI Notetaker | 17 of 19 | 80% | — |
| #5 | tl;dv AI Meeting Recorder | 17 of 19 | 77% | — |
| #6 | Fellow | 14 of 19 | 71% | — |
| #7 | Read AI | 17 of 19 | 68% | — |
| #8 | Avoma | 15 of 19 | 68% | — |
| #9 | Krisp | 13 of 19 | 60% | — |
| #10 | Zoom AI Companion | 11 of 19 | 55% | — |
Add our AI consensus widget to your blog or website — just copy the code below.