Listenr vs notetaker bots
Use a notetaker bot if you need meetings recorded while you are elsewhere, or want everyone on the call to see that recording started. Otherwise, choose on where transcription runs: a bot receives the call inside its vendor’s cloud whatever you set, and most bot-free tools upload the audio too, while Listenr transcribes on your Mac. Bot-free capture alone decides little, because Granola, Fathom, Otter, Read.ai, tl;dv, Fireflies, Notion, Circleback and Krisp all offer it.
| Feature | Listenr | Notetaker bots |
|---|---|---|
| Does a bot join the call | No, capture is from the Mac | Yes, as a participant in the call |
| Meetings you are not in | Cannot record them | Records without you present |
| What other participants see | Nothing, telling people is on you | It announces itself, and Read.ai lets anyone remove it |
| Where your audio is transcribed | On the Mac by default; four of five backends are offline | In the vendor’s cloud |
| Where the recording ends up | Deleted after transcription unless you keep it; kept files stay on the Mac | In the vendor’s cloud, always |
| Who writes the notes | A local model, Apple Intelligence, your own Anthropic or OpenAI key, or Listenr Cloud | The vendor picks; no mainstream bot offers a choice |
| Devices you can use it on | macOS 15 or later, Apple Silicon | Server side, so your own device does not matter |
| What the free plan gives you | Recording, transcription and notes from a local model or your own key; personal use only | Varies widely; Fathom Free has no usage cap |
| Price | Pro $19/month, or $15/month billed yearly | Read.ai Pro $19.75/month, $15 annual; Fireflies Pro $10 annual |
| Training and retention | Local models send nothing; Listenr Cloud stores no transcript | Fireflies contracts 0-day vendor retention; Otter trains on transcriptions |
| Getting at meetings from a terminal or an assistant | CLI and local MCP server on every plan, read-only by default | Fireflies gives every user a GraphQL API; Otter and Granola offer MCP |
“No bot joins your call” used to set a tool apart. Now Granola, Fathom, Otter, Read.ai, tl;dv, Fireflies, Notion, Circleback and Krisp all have a device-capture mode, and Zoom, Teams and Google Meet never needed a bot because the assistant is inside the client. Ask where transcription runs instead; few tools are left.
Where transcription runs
A notetaker bot dials into the meeting from the vendor’s servers, so the audio arrives in that vendor’s infrastructure before anything is transcribed. There is no setting in which the recording exists only on your machine. That says nothing about how well a given vendor handles it.
Most tools that record from the device still send the audio away to be transcribed. Granola’s documentation says it passes audio directly from your microphone and system audio to its transcription provider. Notion says the audio goes to its subprocessors, and names six. Otter describes no on-device processing.
Listenr transcribes on the Mac, with Apple’s on-device engine by default. Four of its five backends run on the device; the fifth sends audio to OpenAI, and only if you select it. No meeting reaches Listenr’s server unless you share it or let Listenr Cloud write the notes, and shared copies never include audio. What leaves your Mac has the detail.
What a bot does that your Mac cannot
The meeting you are not in. A bot joins on a calendar invitation, so a sales lead can review a call an account executive ran alone. Listenr records what is audible on your Mac, so you have to be there.
Consent. A bot is visible. Read.ai’s has to be added by someone invited to the meeting, asks permission to join, shows that it has joined, and leaves when anyone types “opt out” in the chat. Device capture has no equivalent.
Any device. A colleague on Windows or someone dialling in by phone makes no difference to a bot. Listenr needs an Apple Silicon Mac with macOS 15 or later.
Who tells the other people in the call?
You do. When nothing joins the call, the other participants get no signal. Listenr does not announce itself; the only indicators are macOS’s own, on your Mac. The rules on telling people differ by country and by state, and some places require everyone’s consent. If you would rather the tool handled that, a bot is the better fit.
What the cloud gives you in return
Cloud policies differ, so read the specific one. Fireflies states a 0-day data retention policy with its vendors and says user data is never used to train AI models. Otter’s privacy policy, effective 16 June 2026, describes training on de-identified audio and on transcriptions, which may contain personal information.
A cloud service also brings an enforceable retention policy, an audit trail, access from any device and a support line. Listenr has none of that: its library is a SQLite database and files on one Mac, exportable to Markdown and searchable from a terminal. The pricing page has the plans.
Pick Listenr when
The recording and the transcript should stay on your machine, you want to choose the model that writes the notes, and you are on an Apple Silicon Mac. Telling the other participants is then your job.
Pick Notetaker bots when
Take a bot if you need meetings you do not attend, which device capture cannot do, or if every participant should see that recording started. Read.ai’s bot needs host approval, shows that it has joined, and leaves when anyone types opt out in the chat.
Common questions
Which AI notetaker works without a bot?
Most of them now. Granola, Fathom, Otter, Read.ai, tl;dv, Fireflies, Notion, Circleback and Krisp all offer a mode that records from your computer with no bot joining, and Zoom, Teams and Google Meet have assistants built into the client. Listenr records without a bot and also transcribes on your Mac by default, which most bot-free tools do not.
Why does it matter whether a bot joins if the audio is uploaded anyway?
Because a bot removes the choice. A bot receives the call inside its vendor’s infrastructure, so a cloud copy of the audio exists and no setting changes that. Device capture only makes local processing possible, and most device-capture tools still upload the audio to transcribe it. The question to ask is where transcription runs, not where capture happens.
Can Listenr record a meeting I am not in?
No. Listenr records what is audible on your own Mac, so you have to be in the call. This is the one thing a notetaker bot does that device capture cannot, and if you need it, a bot is the right tool. Fireflies, Otter, tl;dv and Read.ai all offer a bot that joins the call.
Do the other participants know Listenr is recording?
Not from Listenr. Nothing joins the call, so there is no participant to see and no recording banner; the only indicators are macOS’s own, on the recording user’s Mac. Telling people is on you, and the rules differ by country and by state, with some places requiring everyone on the call to consent.
Which notetaker bot is the most transparent about consent?
Read.ai’s bot, on its published behaviour. It has to be added by someone invited to the meeting, asks permission to join, shows that it has joined, and leaves when any participant types opt out in the chat. Read.ai’s desktop app can also record without a bot, and then nobody on the call is told.
Try it against your own meetings.
The free plan records, transcribes and writes notes with a model on your Mac. No account needed to find out whether the transcript is good enough.