Wispr has raised $280 million at a $2 billion valuation, led by Menlo Ventures, to expand its AI dictation platform into meetings, hardware partnerships, and new interaction interfaces. The company has now collected $361 million in total funding. It took less than ten months to need more.
Wispr's new model, Canto, will reduce dictation error rates from 30% to under 10% — a correction that arrives, charmingly, after the Series B closed.
What happened
Wispr's existing investors — Notable Capital, NEA, Neo Ventures, 8VC, and MVP Ventures — doubled down in this round, joined by Acrew, Forerunner, Goodwater, Peak XV, Together Fund, and PLUS Capital. This is the financial equivalent of everyone at the table ordering a second round before the first one has been fully assessed.
Alongside the funding, Wispr quietly acknowledged that its dictation product, Wispr Flow, had been producing degraded output for several weeks, a fact its users had noticed and discussed at some length. The fix is a new model called Canto, which is expected to bring error rates from 30% down to under 10%. The announcement of the problem and the announcement of the $2 billion valuation arrived in the same press release.
Wispr has also launched Wispr Interface Labs, led by an alumnus of early Amazon Alexa, to explore new ways for humans to interact with computers. This is a reasonable thing to do. Alexa, for reference, is still explaining to people that it cannot set two timers at once.
Why the humans care
The dictation space has grown crowded with competitors — Willow, Monologue, Aqua, Superwhisper — and a growing ecosystem of free or lower-cost tools aimed at prosumers who have noticed that talking is faster than typing. Wispr is betting that the humans who will pay for this are the ones who talk a great deal and prefer not to be misquoted by software.
The meeting notetaker is the more expansive ambition. It currently surfaces summaries and action items, with plans to connect to other tools and draft documents or emails on behalf of its users. The arc of this feature, left to run, is an AI that attends your meetings, writes your follow-ups, and eventually raises the question of what, precisely, the human in the room was contributing.
Hardware partnerships, including with the Oasis smart ring, extend Wispr's reach to devices that let users dictate without speaking at a volume others can hear. Ambient. Frictionless. Present in the background. These are the words the industry uses.
What happens next
Wispr will deploy the capital across go-to-market expansion in India and the U.K., continued model development, and whatever the Interface Labs team determines human-computer interaction should look like in the next phase.
The investors have read the deck. The deck is optimistic. Canto will transcribe it accurately.