I Tried 7 Meeting Notes Apps. Here's What I Learned.

Search "best meeting notes app" and the SERP looks the way every product-category SERP looks in 2026: ten listicles that rank by vibes, never agree on a winner, and never explain why the apps in question are even the same kind of thing. They are not. Meeting notes apps split into two architectures, and most roundups blur the split so they can fit more affiliate links into the page.

The split is this. Some meeting notes apps are AI transcription bots: they join the call, record everything, and hand you a transcript and an action-item list after. Some are manual surfaces: you type into them during the call, and the app gets out of the way. A few sit in the middle. The "best" app depends on which mode of meeting work you are actually doing, and pretending that question has one answer is the original sin of this genre.

I scored seven meeting notes apps against the same seven-criterion rubric, named where each one wins, and named where each one loses. One disclosure up front: I built one of the seven apps in this list (notetime). I'll get to it last, on the same rubric the others get.

I built notetime because I just wanted to know what I did and when I did it. Most note-taking apps I tried were trying to solve my entire life, which is something that cannot be done, at least not when your meeting calendar is dense enough that the AI summary becomes its own backlog. The thing I actually needed was much smaller: a single surface that timestamps every line and gets out of the way during the meeting itself.

Key Takeaways

• There is no single "best" meeting notes app. The right answer depends on whether you want a bot to attend the call, a manual surface you type into, or a hybrid that does both.
• The rubric is seven axes: capture mode, action-item extraction, live time-anchoring, privacy posture, integration depth, pricing model, and best fit.
• Granola is the strongest AI hybrid where you type and AI fills the gaps (Granola, 2026). Otter.ai is the category-default transcription bot (Otter, 2026). Fireflies.ai is the bot with the deepest CRM and integration surface (Fireflies, 2026). tl;dv is the free-tier-heavy bot alternative (tl;dv, 2026). Notion is the manual surface with AI Meeting Notes available on Business and Enterprise plans only (Notion, 2026). Apple Notes is now the default frictionless pick on Apple devices with native audio recording and live transcription on macOS 26 (Apple, 2026). notetime is the lowest-friction line-level timestamped manual surface.
• The wedge most meeting notes apps miss: when the question you will later ask is "when did we agree to that," an AI summary that flattens the timeline is the wrong tool, and a line-level timestamped log is the right one. The argument for time as the organizing axis sits underneath this whole rubric.

What to Look For in a Meeting Notes App

Most "best meeting notes app" reviews compare on the wrong axes. Number of integrations, sticker libraries, AI screenshot quality. The right axes are the seven that decide whether the app will be useful for the meeting work you actually do, not the meeting work the marketing page is pitched at. The rubric below applies to all seven apps, including the one I built.

The seven criteria

1. Capture mode. Bot (joins the call and records), manual (you type during the meeting), or hybrid (a manual surface that listens in the background and fills gaps). This is the load-bearing axis; the other six are downstream of it.

2. Action-item extraction. Automatic (the app produces a list of action items from the transcript) or manual (you type them yourself, or mark them inline). Higher automation cuts post-meeting work; lower automation keeps the action list honest, since AI-generated action items can confidently invent commitments nobody actually made.

3. Live time-anchoring. Does every line of the record carry the moment it was written or said? Bot transcripts time-anchor by default at the line level. Manual notes mostly do not. When the followup question is "when in the call did we agree to that," line-level timestamps are the answer; an end-of-call summary is not.

4. Privacy posture. Where does the audio go, where does the transcript go, who summarizes it, and is any of it on-device. Confidential, NDA-bound, or one-on-one meetings raise the stakes; "we encrypt at rest" is not the same as "the recording never leaves the device."

5. Integration depth. Calendar auto-join, Zoom/Meet/Teams bot support, Slack, Notion, CRM (Salesforce, HubSpot, Attio), and the long tail. For sales and customer-success teams the CRM integration is the whole product; for solo knowledge workers it is noise.

6. Pricing model. Free, freemium, per-seat, per-org, one-time, BYO. The economics matter most when the app gets deployed at team scale, where a $19/user/month tool meets a 40-person team and becomes a $9,000/year line item.

7. Best fit and honest limitation. Every app wins one cell of the rubric and loses one cell. Naming both is the credibility move.

The rubric is binary or low / mid / high per cell, not a numerical score. Numeric scores look fake and invite challenge. The point is the consistency of applying these seven axes to all seven apps, including the one I built. Every app is going to lose to one of the others on at least one criterion. That is the credibility move. If a row is all highs, the rubric is broken or the post is.

Granola: the AI hybrid where you type and AI listens

Granola is the strongest hybrid in this list, and it is the one app that explicitly does not join your meetings as a bot. You take rough notes during the call (or type one word, or nothing), Granola transcribes the call audio in the background and uses your notes plus the transcript to produce a polished writeup after. Granola positions itself as "the AI Notepad for back-to-back meetings" (Granola, 2026), and the homepage lists Vercel, Linear, Ramp, Brex, Intercom, PostHog, and Replit as customers. The company has raised approximately $125M as of late 2025 (Granola, 2026). On the privacy posture: Granola says it does not retain audio recordings after transcription and does not allow third-party model providers like OpenAI or Anthropic to train on user data (Granola privacy policy, 2026); transcripts themselves live in vendor cloud encrypted at rest and in transit, not end-to-end-encrypted.

Granola at a glance

• Pricing model. Free tier with limited meeting history. Business is $14/user/month, Enterprise is $35/user/month (Granola, 2026).
• Platforms. Mac (primary), iPhone. Windows and web availability not advertised on the homepage.
• Core feature. You type rough notes during the call, Granola transcribes the call audio in the background, and the AI merges your notes plus the transcript into a polished writeup after.
• Best fit. Knowledge workers in back-to-back meetings who want to take notes themselves (for the cognitive value of paying attention) and have the AI fill the gaps without a bot joining the call.
• Honest limitation. Mac-first; the iPhone app exists but the wedge is the desktop experience. The hybrid model is brilliant when the meeting is on your computer and bad when it is not. The free tier limits meeting history, so heavy users push to paid quickly.

Capture mode: hybrid / Action-items: high / Time-anchoring: medium (timestamped transcript, not line-level on your typed notes) / Privacy: medium (vendor cloud; audio not retained post-transcription; no third-party LLM training on user data) / Integrations: high (Slack, Notion, Attio, HubSpot, Affinity, Zapier) / Pricing: per-seat / Best fit: hybrid notetakers on Mac.

Granola wins where the friction is "I want to take notes myself but I cannot keep up." It loses where the friction is "I cannot have a transcription pipeline touch this conversation at all," which is the cell every privacy-bound or in-person meeting falls into.

Otter.ai: the category-default transcription bot

Otter is what most people mean when they say "meeting notes app." It is the longest-running standalone transcription bot, it joins Zoom, Google Meet, and Microsoft Teams calls as OtterPilot, and it produces a transcript plus an AI summary plus action items after each call. It is what comes up first when someone in your company asks for a meeting notes tool.

Otter.ai at a glance

• Pricing model. Basic is free (300 transcription minutes per month, 30-minute max per conversation). Pro is $16.99/user/month or $8.33/user/month annually (1,200 in-app recording minutes, 90-minute max). Business is $30/user/month or $19.99/user/month annually (unlimited minutes, 4-hour max). Enterprise is custom (Otter, 2026).
• Platforms. iOS, Android, web, Mac and Windows desktop, Chrome extension.
• Core feature. OtterPilot auto-joins scheduled meetings on Zoom, Google Meet, and Microsoft Teams, records and transcribes, and produces summaries and action items.
• Best fit. Teams who want a category-default bot that just works on the three major video platforms, who do not need deep CRM workflows, and whose privacy bar is "vendor-cloud is fine."
• Honest limitation. The free tier caps at 300 minutes per month with a 30-minute hard cap per conversation, which is too tight for most regular calendars. The action-item extraction is often confident-but-wrong on calls with ambiguous commitments. The bot in the meeting is itself a presence; one-on-ones change when a third party is recording.

Capture mode: bot / Action-items: high / Time-anchoring: high (line-level transcript timestamps) / Privacy: medium (vendor cloud) / Integrations: high (Zoom, Meet, Teams, Slack, Salesforce, HubSpot) / Pricing: per-seat / Best fit: teams who want a category-default bot.

Otter wins where the friction is "I want a transcript and a summary, every time, without thinking about it." It loses where the friction is "I do not want a bot in the call," which is most one-on-ones, all in-person meetings, and every confidential conversation.

Fireflies.ai: the bot with the deepest integration surface

Fireflies is the transcription bot built for sales and customer-success teams. The product is similar to Otter at the core (joins Zoom, Meet, Teams, transcribes, summarizes), but the wedge is the integration depth and the conversation-intelligence layer: talk-time analytics, AI scoring of sales calls, and CRM auto-sync are the differentiators. Fireflies advertises "unlimited integrations" across CRMs and workflow tools (Fireflies, 2026).

Fireflies.ai at a glance

• Pricing model. Free tier with 800 minutes of storage per seat and limited summaries. Pro is $18/user/month or $10/user/month annually (8,000 minutes/seat storage, unlimited summaries). Business is $29/user/month or $19/user/month annually (unlimited storage, video recording, team analytics). Enterprise is $39/user/month with SSO, SCIM, HIPAA (Fireflies, 2026).
• Platforms. iOS, Android, web, Chrome extension, desktop app. Supports Zoom, Google Meet, Microsoft Teams, and 10+ other conferencing platforms.
• Core feature. Auto-joining transcription bot plus conversation-intelligence layer (AskFred AI assistant, talk-time analytics, AI Skills, Voice Agents).
• Best fit. Sales and customer-success teams who want CRM sync (Salesforce, HubSpot) and conversation analytics, plus teams who care about HIPAA and SSO.
• Honest limitation. The product surface is broader than what most non-sales teams need, and the per-seat cost compounds quickly at team scale. The AskFred AI features and conversation intelligence are aimed at managers more than individual contributors; if you just want a transcript, simpler tools fit better.

Capture mode: bot / Action-items: high / Time-anchoring: high (line-level transcript) / Privacy: medium (vendor cloud; HIPAA on Enterprise) / Integrations: high (CRM-first, Salesforce/HubSpot, Slack, Zapier) / Pricing: per-seat / Best fit: sales and customer-success teams.

Fireflies wins where the friction is "every call needs to land in the CRM by itself." It loses where the friction is "I just want a transcript of my one-on-ones," because the conversation-intelligence layer is mostly noise for that use case.

tl;dv: the free-tier-heavy bot alternative

tl;dv ("too long, didn't view") is the budget-friendly transcription bot. The wedge is an unusually substantive free tier: unlimited recordings and transcriptions in 30+ languages, no credit card, AI summaries included (tl;dv, 2026). The marketing positions it explicitly against Gong's "$5,000+ upfront investment" with "unlimited meetings, AI-generated notes, and multi-meeting summaries at no cost" (tl;dv, 2026).

tl;dv at a glance

• Pricing model. Free Forever plan with unlimited recordings and transcriptions in 30+ languages, no credit card. Paid tier details are not displayed on the homepage; verify pricing in-app before subscribing (tl;dv, 2026).
• Platforms. Web app, supports Zoom, Google Meet, Microsoft Teams. Integrations include HubSpot, Slack, Notion, and roughly 5,000 apps via workflow tools.
• Core feature. Auto-joining transcription bot with AI-generated meeting summaries and multi-meeting summaries (analyze a batch of calls together).
• Best fit. Individuals and small teams who want bot-driven transcription without the per-seat license cost, especially for sales discovery and customer interviews where pulling patterns across multiple calls matters.
• Honest limitation. Pricing transparency on the homepage is limited; paid tier features and prices are not displayed without clicking through. The brand positioning is sales-first ("sales coaching and intelligence platform"), so the product is opinionated toward sales workflows even when the free tier is general-purpose.

Capture mode: bot / Action-items: high / Time-anchoring: high (line-level transcript) / Privacy: medium (vendor cloud) / Integrations: high (Zoom, Meet, Teams, HubSpot, Slack, Notion) / Pricing: free-heavy freemium / Best fit: free-tier-first individual and small-team use.

tl;dv wins where the friction is "I do not want to pay per seat for transcription on a small team or as a solo operator." It loses where the friction is "I need predictable pricing and a clear paid-tier feature matrix on day one."

Notion: the manual surface with AI Meeting Notes

Notion is the meeting notes app for teams that already live in Notion. It is a manual surface (you type during the call), but the 2026 generation of Notion AI now includes an "AI Meeting Notes" feature that "automatically transcribes your meetings, along with a helpful summary" (Notion, 2026). It is bundled into the Business and Enterprise tiers (listed as "Meeting notes Beta" in Notion's pricing table) and is not available on the Free or Plus plans (Notion AI Meeting Notes help, 2026).

Notion at a glance

• Pricing model. Free for individuals. Plus is $10/user/month. Business is $20/user/month. Enterprise is custom. Save up to 20% with annual billing. Notion AI is bundled across Plus and above with limited credits on lower tiers; AI Meeting Notes specifically is gated to Business and Enterprise and is not available on Free or Plus (Notion AI Meeting Notes help, 2026). Custom Agents use a separate credit system ($10 per 1,000 monthly Notion credits) (Notion, 2026).
• Platforms. Web, Mac, Windows, iOS, Android.
• Core feature. Manual notes surface with linkable databases, plus AI Meeting Notes that transcribe and summarize meetings inside a Notion page.
• Best fit. Teams whose knowledge base already lives in Notion, who want meeting notes to land in the same surface as PRDs, roadmaps, and project docs, with the same internal-linking graph.
• Honest limitation. AI Meeting Notes is gated to Business at $20/user/month and Enterprise, which is steep if meeting transcription is the only AI feature you need. The base manual surface is excellent for typed notes but does not auto-timestamp at the line level by default; you get block-level metadata, not real line-by-line timestamps.

Capture mode: manual + AI hybrid / Action-items: medium (manual unless you use AI summaries) / Time-anchoring: medium (block-level, not line-level) / Privacy: medium (vendor cloud) / Integrations: high (Slack, Drive, GitHub, Jira, plus internal Notion graph) / Pricing: per-seat / Best fit: teams who already live in Notion.

Notion wins where the friction is "meeting notes should live in the same place as everything else we write." It loses where the friction is "I want to drop a line into the record in two seconds without picking a database, a template, or a parent page first."

Apple Notes: the default frictionless pick

Apple Notes is the default manual surface on Apple devices, and as of macOS 26 (Tahoe) it now supports "audio sessions right in your note" with "live audio transcripts" that are searchable (Apple, 2026). That feature changes the calculation: Apple Notes is no longer just a typing surface, it is a manual surface with native on-device recording and transcription for the meetings you take on a Mac. If you are an iPhone and Mac user, the meeting-notes question has a stronger default answer in 2026 than it did in 2024.

Apple Notes at a glance

• Pricing model. Free, bundled with iOS, iPadOS, and macOS. No subscription. iCloud sync uses your existing iCloud storage.
• Platforms. iOS, iPadOS, macOS, web (iCloud.com). No Android, no Windows.
• Core feature. Manual notes surface with folders, tags, smart folders, and (on macOS 26) native audio recording with live transcription inside a note.
• Best fit. Apple-only users who want a free, native, zero-install meeting-notes surface; meetings taken on a Mac where the audio-recording-plus-transcript feature is now native; users for whom "no new app to install" is the wedge.
• Honest limitation. Apple-only; no Android, no Windows. Sync uses standard iCloud (end-to-end encryption only with Advanced Data Protection enabled). No bot for Zoom/Meet/Teams; the audio capture is what your microphone hears in the room or on the call. Per-line timestamps are not first-class; the unit of metadata is the note, not the line.

Capture mode: manual + on-device audio (macOS 26+) / Action-items: low (manual) / Time-anchoring: medium (note-level; transcript timestamps on audio) / Privacy: high (on-device or iCloud; ADP for E2EE) / Integrations: low (deep within Apple ecosystem, sparse outside) / Pricing: free / Best fit: Apple-only users, especially those taking notes on a Mac.

Apple Notes wins where the friction is "I do not want to install or pay for anything, and my meetings happen on my own devices." It loses where the friction is "we run on Slack, Zoom, HubSpot, and our notes need to land in the team's shared system." For the deeper head-to-head with notetime see notetime vs Apple Notes.

notetime: the line-level timestamped manual surface

notetime is the app I built. It is not a transcription bot. If your meeting work is "I am on a Zoom and I want the conversation transcribed and summarized for me," every other app in this roundup is built for that and notetime is not. notetime wins in a narrower but real cell: meetings you take notes in by typing, where the question you will later ask is when something was said, not what was said overall.

The cells where notetime is the right answer:

• In-person meetings. Standups, hallway conversations, retros, 1:1s held in a coffee shop. A bot cannot attend; a typing surface with line-level timestamps captures the part you care about.
• Confidential and NDA-bound meetings. Legal review, M&A discussions, board prep, performance conversations, anything where a transcript persisted to vendor cloud is the wrong outcome.
• One-on-ones. The bot in the room changes the conversation. Manual notes do not.
• Standups. The deliverable is "what did I commit to and when," not a transcript of the standup.
• Any meeting where the followup question is timestamp-based. "When did we agree to X" is a question line-level timestamps answer cleanly; an AI summary flattening 45 minutes of conversation into three bullet points does not.

The cells where notetime is not the right answer:

• Long all-hands or town halls you want summarized and want to share with people who missed it. Otter, Fireflies, tl;dv, or Granola fit.
• Sales calls you want auto-synced to CRM. Fireflies or Granola fit.
• Meetings where the action-item list is the deliverable, not the timeline. Any bot fits; the AI summary is the product.

notetime at a glance

• Pricing model. Free for the core feature, no signup. Pro adds cross-device sync, cloud backup, and additional themes at $2.99/month, $29.99/year, or $39.99 lifetime (notetime, 2026).
• Platforms. iOS and web today. iPad and macOS in progress; Android on the roadmap.
• Core feature. Every line you type is auto-timestamped the moment you start typing. One continuous diary. No folders, no tags, no signup. App Store rating sits at 4.9 across the iOS listing (notetime, 2026).
• Best fit. Users whose meeting work is typed notes where the when matters as much as the what, especially in-person and confidential meetings where a bot is not an option.
• Honest limitation. No transcription. No bot. No AI summary. No CRM integration. Text only. If your wedge is "I want a robot to do this for me," notetime is the wrong tool and one of the six apps above is the right one.

Capture mode: manual / Action-items: low (manual, marked inline) / Time-anchoring: high (line-level, automatic) / Privacy: high (offline-first, device-level) / Integrations: low / Pricing: free (Pro is one-time or low subscription) / Best fit: in-person, confidential, and timestamp-driven meetings.

The methodology that sits underneath notetime is append-only journaling, and the cognitive-science case for time as the index is why timestamps beat tags. For the kind of ADHD reader who runs back-to-back meetings, interstitial journaling for ADHD is the closest practice; the two-week note review ritual is what to do with these notes after the meeting.

The honesty I owe the rest of this post: I built notetime for one specific friction (capture cost plus auto-timestamping in meetings where a bot does not fit) and it does not solve the others. The cell where it loses is automation. It is not a bot, not a CRM sink, not an AI summarizer. If your friction is one of those things, one of the six apps above is the right answer.

The Comparison Matrix

The matrix is the at-a-glance view. The per-app sections above are where the nuance lives. If you are skimming, scan for the row that is mostly highs on the criteria you care most about. If you are torn between two apps, read those two sections in full; the matrix is too compressed to make a final decision from.

App	Capture mode	Action-items	Time-anchoring	Privacy	Integrations	Pricing
Granola	hybrid	high	medium (transcript timestamps)	medium	high (Slack, Notion, Attio, HubSpot)	$14 to $35 per seat / month
Otter.ai	bot	high	high (line-level transcript)	medium	high (Zoom, Meet, Teams, Slack, CRM)	$0 / $8.33 / $19.99 per seat (annual)
Fireflies.ai	bot	high	high (line-level transcript)	medium (HIPAA on Enterprise)	high (CRM-first, Salesforce, HubSpot)	$0 / $10 / $19 / $39 per seat (annual)
tl;dv	bot	high	high (line-level transcript)	medium	high (Zoom, Meet, Teams, HubSpot, Slack)	Free Forever; paid tier undisclosed
Notion	manual + AI	medium	medium (block-level)	medium	high (Slack, Drive, GitHub, Jira)	$0 / $10 / $20 per seat (annual)
Apple Notes	manual + on-device audio	low	medium (note-level, transcript on audio)	high (on-device; ADP for E2EE)	low (Apple ecosystem)	free
notetime	manual	low (manual, inline)	high (line-level, automatic)	high (offline-first, device-level)	low	free; Pro one-time or low subscription

No row is all highs. That is the point. Every app in this matrix loses to at least one other on at least one criterion that matters to some user. The right app is the one whose row matches your particular meeting mode, not the one with the most highs overall.

Which One Is Right for You

Seven routing scenarios. Each names one mode of meeting work and the app that fits it best. Read the one that matches you and skip the rest.

You want a bot to attend Zoom / Meet / Teams calls and produce transcripts plus summaries, and you want the category-default option. Otter.ai wins. The free tier is tight; Pro at $8.33/user/month annual is the going rate.

You run a sales or customer-success team and the CRM integration is the whole point. Fireflies.ai wins. The conversation-intelligence layer and CRM auto-sync are the differentiators worth the per-seat cost.

You want a bot but you do not want to pay per seat. tl;dv wins. The Free Forever tier with unlimited recordings is the wedge.

You want to type the notes yourself (because paying attention is the point) but want AI to fill the gaps. Granola wins. The hybrid model is the one architecture nobody else does well, and Mac users get the best of it.

Your team already lives in Notion and you want meeting notes in the same surface as everything else. Notion wins. AI Meeting Notes lands at $20/user/month on Business and is also available on Enterprise; it is not available on Free or Plus.

You are an Apple-only user, your meetings happen on a Mac, and you do not want a new app to install or a subscription to pay. Apple Notes wins. The native audio-plus-live-transcript feature in macOS 26 makes it materially stronger than it was in 2024.

Your meetings are in-person, confidential, or one-on-one, and the question you will later ask is "when did we agree to X." notetime wins. The methodology is append-only journaling; the philosophy is why timestamps beat tags. This is the only place I would steer the reader toward the app I built rather than one of the others.

The honesty here is the point. Saying "Otter wins" or "Fireflies wins" inside a post on this blog is what makes the rest credible. If a comparison post you are reading has the author's product winning every category, the post is a sales page in comparison clothing.

Frequently Asked Questions

What is the best meeting notes app?

There is no single answer; it depends on capture mode. For bot-driven transcription on Zoom/Meet/Teams, Otter.ai is the category default. For sales teams that need CRM auto-sync, Fireflies.ai fits. For a hybrid where you type and AI fills the gaps without a bot in the call, Granola fits. For team knowledge bases already in Notion, Notion's AI Meeting Notes fits. For Apple-only users on Mac with native audio recording, Apple Notes is now the default frictionless pick. For in-person, confidential, or timestamp-driven meetings, notetime fits. Pick the row whose match is closest to your meeting mode.

Are AI meeting notes accurate?

Transcription accuracy is high on clean audio (single language, low background noise, headset mic) and degrades on noisy or multi-speaker calls. AI-generated action items and summaries are confident-but-sometimes-wrong: the model will assert commitments and decisions that did not actually happen, especially on ambiguous discussions. The honest workflow is to treat AI summaries as a starting draft and verify against the transcript, not as a final record.

Is there a free meeting notes app?

Yes. Otter.ai's free Basic plan offers 300 minutes per month with a 30-minute conversation cap. Fireflies.ai's free plan offers 800 minutes of storage per seat. tl;dv has a Free Forever plan with unlimited recordings. Notion's free plan does not include AI Meeting Notes (that feature is on Business and Enterprise only), but the manual notes surface is free. Apple Notes is fully free on iOS and macOS. notetime is free for the core feature with no signup. Free options exist; the question is which free tier's shape matches your mode.

Should I use a transcription bot or take manual notes?

It depends on the meeting. Bots fit recurring video calls where you want a transcript to share and the participants are comfortable with recording. Manual notes (or hybrids like Granola) fit calls where paying attention is the point, in-person meetings, confidential conversations, and one-on-ones where the bot itself would change the dynamic. The same person can use both for different meetings, and most professionals do.

What is the best meeting notes app for in-person meetings?

A manual surface with line-level timestamps. Bots cannot attend an in-person meeting; an end-of-call transcript does not exist. The right tool is one that lowers capture friction during the meeting (so you can keep up with what is being said) and preserves the when of each line afterward. The methodology is append-only journaling; the apps that fit this shape are Apple Notes (no auto-line-stamps) and notetime (auto-line-stamps).

How do I keep meeting notes confidential?

Use a tool that keeps audio and transcript off vendor cloud, or use a manual surface with no recording at all. Apple Notes with Advanced Data Protection enabled, or notetime offline, are the closest to that bar in this list. Otter, Fireflies, tl;dv, and Granola all send audio or transcripts to vendor infrastructure; the privacy posture is "vendor-cloud, encrypted in transit and at rest," not "off-device." For NDA-bound, M&A, or board-level meetings, that distinction matters.

The Short Version

There is no single best meeting notes app. The right one is the one whose capture mode matches the meeting work you actually do, scored against the seven criteria above. Bots fit some meetings; manual surfaces fit others; the hybrid fits a third bucket. Pick the row in the matrix that matches your mode and stop reading.

I built notetime because the friction I hit hardest was the in-person meeting and the confidential one-on-one, where a bot is not an option and a flat AI summary loses the when. That is not the same as saying it is the best of the seven. It is one of them, with a real strength on one cell of the rubric and an honest set of misses on the others.