Wispr raises $280 million and sets its sights on meetings, not just dictation
The AI dictation startup is now valued at $2 billion and just launched a new voice model that it says will cut speech errors by two-thirds.
Key points
- Wispr raised $280 million in Series B funding led by Menlo Ventures, valuing the company at $2 billion as of Monday.
- The startup has now raised $361 million in total, with its previous round closing less than 10 months ago.
- Wispr's new voice model, called Canto, aims to cut dictation error rates from 30% to below 10%.
- The company is expanding into AI meeting notes, competing with tools like Granola and Fireflies.
- Increased competition from free and low-cost dictation apps adds pressure on Wispr to justify its premium.
Wispr makes software that turns spoken words into typed text, letting you dictate emails and documents instead of typing. On Monday the company announced it's raised $280 million in a Series B round, the name for a mid-stage startup funding raise, led by venture firm Menlo Ventures. The deal values Wispr at $2 billion.
That's a lot of money, fast. Wispr's previous round closed less than 10 months ago, and the company has now raised $361 million in total. We first covered Wispr on 17 August 2026, when voice-input tools were already drawing serious investor attention.
What is Wispr actually spending the money on?
Wispr wants to move beyond voice-to-text. The most concrete signal is a new meeting notetaker that produces summaries and flags action items after a call, putting Wispr in direct competition with Granola and Fireflies.
The company also launched Wispr Interface Labs, a research arm led by Ariya Rastrow, one of the early engineers behind Amazon Alexa, to explore what talking to your computer might look like beyond dictation.
On hardware, Wispr is partnering with devices like the Oasis smart ring so users can dictate quietly without speaking aloud.
| What changed | Detail |
|---|---|
| Funding raised | $280 million Series B |
| Total raised to date | $361 million |
| Valuation | $2 billion |
| New model | Canto, targeting under 10% error rate |
| Previous error rate | Around 30% |
| New product | AI meeting notetaker |
Should current users notice a difference?
Yes, and the timing matters. In recent weeks, a number of Wispr users complained publicly that dictation quality had dropped. The company acknowledged the problem and says Canto will fix it, bringing error rates down from roughly 30% to under 10%.
A 30% error rate means roughly one word in three comes out wrong. Getting that under 10% would make the tool genuinely useful for professional writing again.
Wispr has also expanded to Android phones and is growing its sales teams in India and the U.K.
What is the honest catch here?
Competition is sharpening. Apps including Willow, Monologue, Aqua and Superwhisper all do similar things, and several are free or cheaper than Wispr. When rivals charge less and the gap between products narrows, a $2 billion valuation demands a convincing answer: why pay more?
Wispr is betting that bundling dictation, hardware partnerships and meeting notes into one tool justifies the price. That might work. But almost every startup raises a big round with a bold story, and it's worth remembering that most aren't featured in the press releases that follow a few years later.
First reported by TechCrunch AI, the funding confirms investors believe there's serious money in voice-input. Whether Wispr captures it is an open question.
The one thing worth doing now: if you already pay for a dictation tool, check whether a free alternative covers what you need. If it does, the savings are real money in your pocket, not Wispr's investors'.
