a 24 year old led the engineering behind our second product ever. it sounds insane, but it's been one of the best decisions we've made. he was only a couple of years out of college. one day i pulled him aside and told him i needed him to own this product for the next two months. the team working on this started very small. what i needed for notetaker was someone with real intellectual horsepower who was fast at building. i didn't need them to be experienced. he started with small experiments and two months later he had helped design how the product worked, the systems it runs on, and was managing other engineers on the project. that's 2-4 years of promotions at any other tech company, done in eight weeks. the only thing i did was meet with him for 30 minutes a day. young and inexperienced aren't the same thing. someone hungry to grow, sitting next to someone who can pull them forward, moves faster than you'd ever expect. we're hiring people with these qualities now more than ever.

We have a few core beliefs at Wispr that guide everything we do. One of them is that you should be able to dictate wherever you are in the world. That means understanding you in a busy coffee shop, with a conversation happening at the next desk, or when you’re speaking quietly in a shared office space. That's why, today, we're introducing our latest speech model, Canto. It's the first in a rapid series of speech models we’re building to make that belief a reality. We built it specifically around those everyday conditions. On our evaluation of 10 hours of English-language Wispr Flow dictations from more than 2,300 speakers, it achieved a 3.4% word error rate—the lowest among the models we tested. We’re moving into a voice-led future, and I’m happy to share a closer look at the work our team is doing to get us there. Learn more about Canto here: wisprflow.ai/canto

There's always a trade-off that comes in between quality, scope, and time. The way that we run this inside Wispr is: Quality comes first Time comes second Scope comes third If something is not up to the Wispr bar, we just push the deadline further away. We don't ship it if it's not meeting the quality bar. However, what we can do is reduce scope. We can ship a very narrow thing that's very high quality. I would rather do that than ship a lot of features of lower quality. UX always comes first.

Today is an exciting one: Notetaker is now live on Windows. When we launched Notetaker, I wrote about how much we hate messy meeting notes. We wanted something that gets the details right and makes it easy to push your work forward. That becomes especially useful a few meetings later. You’re writing a proposal and need to remember what the customer asked for. You’re planning next week and want to check which decisions are still open. The context you need is spread across several conversations. With Notetaker, you can ask questions across your recorded meetings and get answers with links back to the source. Your personal dictionary helps it recognize the names and terminology you use at work, making those conversations easier to follow when you come back to them. You can also bring that meeting context into your AI tool of choice, alongside the documents you’re already working with. A conversation with a customer can inform a proposal. A discussion with your team can become the starting point for a project plan. So much of our work starts with something someone said. We want that context to stay useful as the work develops, whether you’re following up that afternoon or revisiting a decision weeks later. A good conversation should keep being useful after the meeting ends, and that’s exactly why we’ve built Notetaker.

we always try to ship sooner than something is ready doing this allows us to constantly build alongside real user feedback if you don't do this you can find yourself in this perfection mindset making very marginal improvements, and before you know it, a year has passed and nothing has shipped this is what happens to most startups, and exactly what you want to avoid ship fast, gather feedback, improve, ship again

Some moments make you stop and think "where am I?" I grew up in Delhi dreaming of building technology that millions of people couldn’t live without. Recently, I returned to New Delhi as a speaker at The Economic Times World Leaders Forum, and had the honor of meeting Prime Minister Narendra Modi. On stage, we talked about what it really means to build an AI-native company, and why India’s AI future shouldn’t be a copy of the American playbook. India is a voice-first country. We talk on our phones, send voice notes, and communicate across dozens of languages. But building for India also means designing for exceptional affordability, local context, high expectations, and the realities of people across different cities. If we get that right, voice can make technology easier and more natural to use—giving millions of people a simpler way to communicate, create, and get things done. Thank you to The Economic Times and the ET World Leaders Forum team for bringing together builders, investors, policymakers, and leaders around such important conversations. I’m also grateful to my fellow panelists and everyone who made the event possible. To come home, share what we’re building, and meet the Prime Minister of India was a moment I won’t forget.

We recently launched our second product surface, Notetaker. This recent episode is what 72 hrs before a launch looks like at Wispr - lots of testing, finishing touches, and a team coming together to make something we're proud of. piped.video/channel/UCOMEQcX…

In India, we have a saying: “Rishte lafzon ke mohtaaj nahi hote.” Relationships don’t depend on words. Because in India, sometimes “I love you” sounds like: “Have you forgotten us now that you’re out exploring the world?” or “I hope you’re keeping up with your physiotherapy after the surgery!” We don’t always say exactly what we feel. A best friend pulls you into a hug without explaining why. A partner disguises care as a casual reminder. The words may struggle to come out. But expressing them shouldn’t feel like work. That’s what I love about this film, and about Wispr Flow. It makes communication feel seamless, especially when finding the right words isn’t. — Written with @WisprFlow

Learn more about our Series B: wisprflow.ai/post/series-b

I want to talk about a video we posted last week, the one titled "Google is coming for us." Most of the replies fell into two categories.... 1. "Wispr is a wrapper anyone could build in a weekend." Yes, you can vibe code a version of Wispr. We encourage it, honestly. It's a fun build and you'll learn a lot. But getting from 6/10 to 9/10 is incredibly hard. Making dictation feel instant, getting names and jargon right every time, formatting a text differently in an email, editing mid sentence with just your voice, these are incredibly tough (fun) challenges. Most models cap out well before what top-tier voice dictation needs to be to deserve trusted, habitual usage. We deeply understand this problem - because it is all we have thought about for years - and recently launched Wispr Advanced Interfaces Lab to create a step change in voice quality. We see how this can be significantly better and will prove it. 2. "Wispr has gotten too hype." This post lacked the authenticity we aim to deliver. We started building Wispr in public closely with the X community and we want to make sure we are as grounded and humble as we always have been. We will continue to have fun, be open and push the boundaries through our content, models, and product innovations. -- Going forward, we'll be releasing more benchmarks, regular product updates, challenges we face, and incorporating all the product feedback we get here back into what we're building. The future of voice is an exciting one, and we care about what you think we should do next!

We'd like to clear this up, as privacy is central to everything we do at Wispr: 1) This data is from a small subset of users that have opted into this to improve models. 2) Our onboarding surfaces privacy mode as an option for all users: nothing you say is used for model training. 3) The astroturfing on this platform is getting out of hand

Upgraded to a Wispr Flow mic Rarely use a keyboard anymore

i killed our india ad budget and wrapped 100 autorickshaws in @wisprflow branding, driving laps around bangalore. seven weeks later, one of the fastest-growing companies in the country emailed us. my team thought i'd lost it. rickshaws don't scale, and there's no way to attribute a sale to a vehicle sitting in traffic all day. we ran them anyway. then the email came in. @ZeptoNow's partnerships team had seen the autos and wanted to talk. the smartest thing i did was forward it to our india lead @findingnimo_. seven weeks later, this is what @vasubhagat011 turned it into: - a film where zepto's own people plan their 5th birthday with wispr flow. dictating mid-air, flicking the phone across the conference room, nobody typing. - placements inside their app, on the screen twenty million people watch while they wait for their groceries. - zepto's top 100 leaders working with us for the next year. zepto turns 5 this week, and their birthday got planned the way they run the company. fast, and in flow.

5 years ago this week, we founded Wispr we've made every mistake imaginable. including building a brain-computer interface our own team didn't want to use but at year 3, we pivoted and launched Wispr Flow. now, millions of words flow through it daily. And yesterday, we shipped our second product, Notetaker the culture we've built is kind and aggressive. Our team faces the hardest problems head-on with enthusiasm If you want to see behind-the-scenes of what we're building, our culture, and our team, you can here: linktw.in/JRPxZn here's to 5 more.

We launched Notetaker today, and there's a few key features we're excited for you to try: - Speaker labeling (helps your LLM of choice keep track of action items) - The "What did I miss" button (for the times you need to answer the doorbell and don't want to miss anything) Here's how they work 👇