Wispr Flow built its name on an app that turns speech into typed text, fast and accurate enough that people use it instead of a keyboard. Its new product, Notetaker, moves the same technology into meetings: it listens in, transcribes, and turns the conversation into summaries, decisions, and action items.
The company is not a small player experimenting on the side. It raised 280 million dollars in Series B funding earlier this year, led by Menlo Ventures, at a valuation of 2 billion dollars. That round came less than a year after the company had pivoted away from its original plan entirely.
Wispr actually started in 2021 as a wearable hardware company, building a device meant to capture speech silently. The founders eventually noticed that people preferred the software layer on its own and dropped the hardware altogether, betting the whole company on the dictation app instead. That pivot is now paying off, and Notetaker is the company's first serious move beyond dictation into the crowded meeting notes market.
And it is crowded. Otter, Fireflies, Fathom, Read AI, and Granola are already fighting for the same customers, all offering some version of automatic transcripts, summaries, and searchable meeting history. Wispr's pitch is accuracy: it pulls from your calendar, your personal dictionary of names and terms, and connected tools like Gmail and Slack so that unusual names and company jargon get transcribed correctly the first time.
The more interesting detail is how it records. Instead of joining a call as a visible participant with a name and an icon, the way Otter and Fireflies typically do, Notetaker captures audio directly from your computer in the background. This is the same approach Granola built its reputation on, and it makes meetings feel less cluttered since nobody sees a bot sitting in the call.
But it also means the people you are talking to may have no built-in signal that a recording is happening. Wispr's own product documentation is direct about this: the company states plainly that responsibility for telling everyone on the call falls on the person using the app, not the software.
This is not a small footnote. Otter.ai is currently defending lawsuits that argue the company recorded conversations and used them to train its models without proper consent, and some of these cases are pulling in state biometric privacy laws because notetakers try to identify who is speaking. The entire category of bot-free notetakers is being tested in court right now, and the outcome will shape what every vendor in this space is allowed to do by default.
For any business thinking about rolling out a tool like this, the product quality is only half the decision. The other half is making sure your team actually tells clients, candidates, and partners before they get recorded, in whatever language your local law requires. Wispr does offer enterprise controls to require a consent acknowledgment before recording starts, but that feature has to be turned on. It is not automatic.