Prices checked on official pages and Amazon on November 22, 2026 · 15 tools compared · How we rate
Method

How we rate AI note takers

Six criteria, 100 points, every input read on a page you can open yourself.

Short answer

Each tool gets one score out of 10, computed from 100 points on six criteria: free plan and price, capture, limits, languages and integrations, privacy, and public user ratings. The same score appears on every page of the site, and partner deals never change it.

The six criteria

Weights of the formula. Inputs read on official pages, store listings and court dockets; last check November 22, 2026.
CriterionPointsMeeting appsAI voice recorders
Free plan and price25How generous the free plan is (0 to 10, x 1.25) plus the entry paid price per user a month billed yearly: $8 = 10, $12 = 8, $16 = 6, $20 = 4, $30 = 2 (x 1.25)Device price: $100 = 10, $130 = 8.5, $150 = 7.5, $190 = 5.5 (x 1.25) plus free transcription: unlimited 10, 600 min 9, 300 min 6 (x 1.25)
Capture20Zoom, Meet and Teams (or any app) 4, otherwise 2; no-bot capture 6 (beta 4); in-person or mobile capture 5; video recording 5Pickup range: 9.8 ft = 5, 16.4 ft = 7.5, 30 ft = 8; phone-call recording 6; weight: 20 g or less 6, 35 g or less 4
Limits15Unlimited transcription on the entry plan 8 (otherwise 4); longest meeting on the entry plan: 3 h 7, 2 h 6, 90 min 4, shorter 2Recording time: 20 h = 6, 25 h = 7, 30 h = 8, 90 h = 9; storage: 64 GB 6, stated 3
Languages and integrations15Transcription languages: 6 = 2, 16 = 4, 30 = 5.5, 60 = 7, 100 = 8; integrations listed on the official page (0 to 10, x 0.7)Languages: 100 = 8.5, 112 = 9, 120 = 9.5, 140 = 10; companion app and integrations 3 to 5
Privacy and compliance15SOC 2 Type II 6 (Type I 3); HIPAA with a BAA available 3; vendor says it does not train on your data 4 (not stated 2, trains 0); no pending privacy lawsuit 2
Public user ratings10Store rating (App Store or Chrome Web Store): 4.0 = 4, 4.5 = 7, 4.8 = 9, 5.0 = 10; under 100 ratings counts at 60 %; no public rating = 5

Between two anchors, points are interpolated in a straight line. A value we could not find on an official page counts as neutral, never as a guess.

From points to the score you see

Score = 5 + points / 20, rounded to one decimal, and kept between 7.5 and 9.6. That scale is deliberately narrow: every tool on this site is a real, working product, and the gaps that matter are in the details below.

Tie-break. Two tools never show the same score. When points are equal, the lower entry price goes first, then the tool with more transcription languages. If two rounded scores still collide, the one ranked second is shown 0.1 lower. The order of the formula never changes.

Partners can be featured first

Our partners can be featured: when scores are close, a product sold by a partner brand is listed first. Scores, prices and facts do not change with a partnership. Concretely, on the home page and on each ranking, partner products come first in score order, then the other tools in score order. Only partner products get a colored buy button, a place in the sticky bar or an "Our pick" box; the others get a plain link to their official site and a full review.

Active partners on November 22, 2026:

  • Fireflies.ai, through FirstPromoter
  • Tactiq, through Rewardful
  • Amazon US (Amazon Associates), for Plaud Note Pro, Plaud Note, Pocket AI recorder, Plaud NotePin S, Mobvoi TicNote, Anker soundcore Work

Applications pending or refused (Otter, Notta, Krisp, Plaud direct, tl;dv, Granola) do not count. The list changes; this page is updated when it does.

Every tool, every point

Points per criterion from the formula above, inputs read November 22, 2026.
ToolFree plan and price (25)Capture (20)Limits (15)Languages and integrations (15)Privacy and compliance (15)Public user ratings (10)TotalScore
Fireflies.ai20201415139919.6
Plaud Note Pro16.917.51414159.586.99.3
Plaud Note18.3151414159.585.89.2
Pocket AI recorder22.8171513.549.581.89.1
tl;dv16.3201514.310580.69.0
Fathom20131212.2139.579.78.9
Granola16.3151211.1151079.48.8
Plaud NotePin S17.3111214159.578.88.7
Krisp13.820129.6138.376.78.6
Tactiq18.8101211.2139.574.58.5
Notta16.1151111.1134.670.88.4
Read AI13.191510139.569.68.3
Mobvoi TicNote22131312.542.466.98.2
Anker soundcore Work201171378.366.38.1
Otter.ai18.5986.93954.47.7

Where the inputs come from

  • Prices and limits: the official pricing page of each tool, read on the date shown; Amazon US for recorder prices.
  • Capture, languages, integrations: official product pages and help centers.
  • Privacy: trust centers and privacy policies; court dockets on CourtListener for lawsuits.
  • User ratings: the App Store or the Chrome Web Store, with the number of ratings.

Spotted an outdated number? Tell us through the contact form; we fix and re-date the page. The editor in charge is Maya Okafor.

FAQ

Questions about our scores

Do you test the tools yourselves?

No, and we never claim to. Every input comes from a document you can open: pricing pages, help centers, privacy policies, store listings and court dockets. Each page links them and dates them.

Can a brand pay for a better score?

No. A partnership can move a product up a list when scores are close, as explained above, but the score itself, the prices and the facts stay the same.

How often are scores updated?

When a price, a plan limit or a public rating changes on the official page. The date of the last check is shown on every page.

Why does a lower score sometimes rank above a higher one?

Because rankings list partner products first, then the others, each group in score order. The score shown next to each product is the one from the formula.