Plaud vs Bee vs Omi vs the Watch You Own

The wearable question only starts when the conversation has no invite. If your important conversations are scheduled video calls, a meeting bot beats every wearable on this page and costs you nothing new to carry. It joins, transcribes both sides cleanly, and posts to Slack. If that is most of your week, buy that instead and stop reading.

The conversations people tell me they lose were never scheduled. The twenty minutes after the meeting, when the real decision gets made. The client who says the important thing on the walk to the lift. An executive assistant put it plainly: "Some meetings are online, but a lot are in person meetings or quick follow ups where setting up a full meeting bot feels weird."

There is a quieter second reason. Your company's recording belongs to your company. If you want your own record of what you agreed to, do not keep it somewhere your employer administers.

Bee and Omi both ship Apple Watch apps of their own, so two of the three pendant makers already let you capture without the pendant. For the in-person capture compared here, Plaud still centers on its own recorder. Its newer Desktop beta also captures online meetings without a bot, so this is a comparison of room-recording setups, not a claim that Plaud records only through hardware. So the hardware line is not where the marketing puts it.

What the hardware buys is immunity to one specific failure. On a watch or a phone the microphone is shared, and the system decides who gets it. A call arrives, iOS hands the audio session to the call, and whatever was recording loses it. A pendant never enters that fight, because its mic sits on a device that is not taking the call. An independent microphone is one reason to buy a pendant; offline storage, placement and battery can matter too. A device that streams through your phone can still lose the stream, so a separate microphone alone does not guarantee an uninterrupted recording.

An app can ask not to be interrupted, and the ask is conditional. Apple's documentation says that preference "has no effect if the device uses the full-screen display style", and a locked iPhone shows every call full-screen whatever you have set. Answer the call and the audio session goes either way. Apple's own Voice Memos does not survive it: "If the other app starts playing audio, Voice Memos stops recording."

So the question is only how often that happens to you. If Plaud is the one you are weighing, the Apple Watch alternative to Plaud has its own decision guide. Here is what the devices cost.

Five devices from four makers are worth a look right now, running from fifty dollars to two hundred and fifty. One spec explains most of that spread: whether the thing works at all with your phone in another room. Subscription prices are the yearly rate, per month.

Bee is the cheapest way in, but it ships only to the United States. Its stream also needs your phone free to receive it, which is what an incoming call puts in doubt. Bee does not document what happens to a stream already running.

The thirty dollars between the NotePin and the Note Pro buys a screen to review on, not more storage. Omi records offline too, but publishes no capacity for the device it actually sells; its developer docs put 8GB on the DevKit hardware, an eighth of Plaud's. Friend publishes nothing at all and calls itself a companion rather than a note taker, so that is the last you will hear of it.

Battery is the other pitch, and a fair one. The Bee, original NotePin and Omi figures are the reported use results carried over from Forbes Vetted’s comparison, not a uniform lab test. Plaud currently advertises up to 50 hours for the Note Pro; the 30-hour figure on the same comparison page belongs to the original Note. Those are manufacturer claims, not our measurements. A Watch has a different routine: it is on your wrist already and on a charger overnight. What matters is whether one long session leaves enough battery for your day. Our earlier Series 9 observation put a day with capture running at roughly 10 to 20 percent extra drain with a reasonably healthy battery. It was not a controlled comparison across apps; hardware, battery health, connectivity and other Watch use change the result.

A pendant can earn its place through microphone placement, battery or independent storage. You have to remember to wear it. There is a second charger, a second thing to lose, and for most of them a subscription on top of what you already paid. Buy hardware for a specific recording problem it solves, and judge the notes separately.

Limitless is not in the table because you cannot buy it. Meta acquired the company, the Pendant stopped selling on 5 December 2025, and existing owners are supported through 2026 without a subscription.

In Brazil, China, the European Union, Israel, South Korea, Turkey and the United Kingdom the service ended that same day. Customers there had until 19 December to download their transcripts, "after which, their account and all associated information will be permanently deleted." Rewind is shutting down too. Nothing records any more, though what was already recorded stays readable through 2026.

Omi publishes a guide to integrating third-party wearables that names Limitless, but as something you wire up yourself. Hardware money goes in up front, and what it is worth in a year depends on decisions you do not get to make.

If you would rather use the Watch you already own, audio quality is not what decides it. Session length, speaker separation and summary timing do, and half the field publishes nothing at all on the first.

Sit with the Bee and Omi rows. Both sell a pendant and also offer Watch capture. That gives you a way to try the software before deciding whether separate hardware solves a real problem. Omi's also streams to a paired iPhone, so it wants the phone in range too.

When the summary appears

What about Siri Recap? Apple’s announced Siri Recap is designed for brief conversation summaries, not full transcripts. Its beta is due later in 2026 on compatible hardware. If you need a searchable transcript and follow-up work now, try Pastok.

• Most of the field waits until you stop. You talk, you stop, it processes. Right shape for one person thinking out loud on a walk. Wrong shape for four people going round in circles for an hour, because nothing exists until it is over.
• Pastok builds it while people are still talking. The transcript comes first, the summary a few minutes behind it. You do not finish a 60-minute meeting and then sit there waiting for it to think. By the time you are shaking hands the structure is already there.

What happens when a call comes in

Every app on a Watch lives with the interruption from earlier, because none of them own the microphone. The best any of them can do is notice fast and get you back. The watch tells you capture has stopped. You bring the app forward and resume. Pastok does that, and it does not get the lost minute back. If you find a Watch product claiming it never gets interrupted, try it with a real incoming call before you believe it.

What differs is recovery. Check whether an interrupted recording can resume, whether it becomes a second file, and whether the app can use both parts afterwards. Pastok can keep earlier processed context available when capture resumes, but it cannot recover audio that was never captured. Granola also offers chat across meetings, so two files do not by themselves prove two disconnected answers.

What speaker labels are worth

Speaker separation is where I have to watch myself, since it is my own product. Pastok labels speakers with anonymous IDs, and overlap or distance can still confuse them. Other products advertise different attribution features, including voice profiles, but a label in a feature list is not a test of your room. Check a short sample with the people you actually meet. Treat automatic labels as a review aid, never as approved minutes.

Every product here leads with the transcript, and almost nobody reads one. A manager sitting through eight meetings a day put it this way: "I either don't read it or ask AI to summarize if I have action items."

So the transcript is not the deliverable. What you want back is the conversation already reduced to decisions, owners and open questions by the time you are at your desk.

• Activity summary and Action Board, both built while the session is still running.
• Chat, to ask questions of it afterwards, or to send off and research something and hold the answer up against the argument you actually had.
• Skills, which run on demand, on a schedule, or when a situation comes up, and come out as Markdown or PDF.

A Skill is a standing instruction you write once: a client-call debrief, say, or a weekly review of what you have committed to. Write it in the morning and it runs on every call you take that month.

Moving that record into whatever else you use is the one thing Granola and Wave do better. Both ship native connectors to Notion, Slack and HubSpot. If something has to land in a CRM with nobody touching it, buy one of those. Ours goes out three ways instead, none of them a connector.

• Paste. An Activity drops straight into Claude Code, Codex or Cursor, with no server and no setup.
• MCP, live and free on every account including ones that have never paid, reaching the last seven calendar days read-only.
• Export, for anything older than those seven days.

Everything is on the web app if you would rather copy it out yourself, and there is a page of setup steps for each client. Someone took a café conversation about a startup idea and had an agent build the demo from it that afternoon. That is a paste, not an integration, and it ends with something existing.

Six months on, none of it matters if you cannot find the thing again. Assume every paid tier keeps your transcripts until you delete them, so storage is not the interesting question. Pastok searches earlier sessions by meaning, so a conversation whose date you have forgotten turns up when you describe what was said in it. It keeps a long-term memory of two kinds: what stays true about you, and the sessions themselves. So a Skill you write today can draw on a conversation from months ago.

Three things decide this and they do not move together. Who owns the vendor now, whether it trains on your data, how long it keeps your raw audio. Two of the six products here changed hands in 2025, and those two acquisitions cut in opposite directions.

Granola does not retain meeting audio after processing; mobile recording uses a temporary cache. Its own training setting and third-party provider rules are separate. Bee stores none either and does not train, and it belongs to Amazon. Friend's marketing and its own privacy policy contradict each other, which tells you something before you have read either. And Limitless is the row where a retention setting stopped meaning anything. Whatever a customer in the EU had picked, the answer became zero on a date Meta chose.

Nobody wins that table.