Get live statistics and analysis of Gregor Zunic's profile on X / Twitter

We passed OpenAI SDK on Github

OpenAI operator is cool. Wouldn’t it be nice if it was open source?👀 We built Browser Use - 100% OSS version you can use today for free Link below ↓

We Raised $17M to Build the Future of Web for Agents 🤖 A few months ago, Browser Use was just an idea weekend experiment to see if LLMs could navigate the web like humans. In just four days, we built an initial prototype and launched it on Hacker News. Build stuff, move fast, break things💻 Today, we are excited to share that we have raised $17 million to take web agents to the next level. This funding round was led by Felicis Ventures, with participation from A Capital, Nexus Ventures, Y Combinator, Paul Graham, Liquid2, SV Angel, Pioneer Fund, and others. Thanks everyone for all the support♥️ ps* we are hiring https://t.co/QmZDo5jQmr

Introducing: Browser Use Cloud☁️ All the features of OpenAI Operator for $30/month. Or you can just use the OSS library completely for free. Your choice. That's the beauty of Open Source♥️ The cloud version offers proxies, persistent authentication, message history, and human in the loop. Oh, and browser repos combined have hit 23k⭐ on Github. Thanks everyone for helping to push this field to the new heights. We will push harder than ever to make the web agents in the world. WebVoyager 95%+ soon. Link to the cloud below ↓

Browser Use seems to have been the fastest growing repo on Github in January🌚

Browser Use is Hiring the top 0.01% Founding Engineer to build internet for LLMs🔥 We (2 people) have built the leading repository for web agents—45K+ GitHub stars in just 4 months. Every day, someone tells us about a wild new use case at some hackathon. But we’re just getting started📈 We just wrapped up fundraising and now we need help. We are MELTING under customer demand and want to push the edge of what’s possible. To do that, we need the most cracked engineers on the planet. 🚀 Who we’re looking for: - Top 0.01% engineers with no life—if your GitHub isn’t a dark green carpet it’s a red flag. - You wake up thinking about code and go to sleep thinking about code. - You grind every day because you want to build something that actually matters. - you like solving hard problems all day long and move very fast (much faster than OpenAI - proof is chart below) 👀 Know someone like this? We’ll pay you if we hire them. 💀 Want work-life balance? Go to Google. 🚀 Want to build the future of web agents and grind hard problems? Come to SF. Both Magnus and I have been grinding startups for years with zero breaks. We only hire software & ML engineers, so the team culture is pure software geekery—no distractions, just cracked builders shipping daily. We did all of this with 2 people. Imagine what we can do with 5. If this is you (or you know someone) dm me. We offer very good equity+salary. We promise that you will never be bored with your work. Let’s build the next web agent unicorn together🦄

So… apparently Manus is just a Browser Use wrapper

Browser Use crossed 35,000 ⭐️ on GitHub Wanted to take this surreal opportunity to explain how all of this madness got started. After leaving our previous startups, Magnus and I found ourselves at ETH Zurich’s Student Project House—an incubator where you’re only allowed to build startups, not study (one of the coolest things ETH is doing, in my opinion). We started talking about how web agents were going to be the big thing in 2025, even predicting that agentic clicks might soon outnumber human clicks on the web🕸️ What started as casual brainstorming over a few lunches turned into a challenge: let’s build something small, throw it on Hacker News, and see what happens. We put together an MVP in four days, launched it on Show HN… and boom—number one. From there, it’s been an absolute rocket🚀 Zero to 35,000 GitHub stars in four months. Completely surreal. What started as a side project has grown into something so much bigger, and taking it from a weekend experiment to a real company has been the most exciting (and challenging) thing we’ve ever done. Feels like this is what we were meant to do🦄

Are you tired of waiting for OpenAI to release the Operator API?⏳ Say no more! Browser Use API is here (link below)🌐 I guess two OSS guys from Europe can ship faster than 100 people from ClosedAI.

50K GitHub ⭐ & the two eras of life: pre-YC vs post-YC 🦸 Yesterday, Browser Use hit 50,000 stars. It’s been less than 5 months since we started, which gave me some time to reflect (also everyone should read the book Shoe Dog). The real story is what happened after things started going viral—how YC changed everything📸. A few weeks after Hacker News blew up, we applied to YC and got an interview. Then: total silence for 3 weeks🤫 We didn’t sit around. We pushed like hell. Every week we sent updates to YC: • People using us in production • Real customers • Equity hire offers • Sales convos with big tech companies Still nothing. But we didn’t stop. We were 100% sure we were getting in. Then, finally: re-interview → got in. Coming from Zurich to SF was rough. I hated it at first. But the YC community flipped everything. You’re suddenly surrounded by the most cracked founders alive. Your ambition 100x’s overnight. Before YC, I thought building a unicorn was one step: → get into YC → build a unicorn Turns out there’s a milestone every month. It’s way harder. But it’s also way more fun. YC created a clear before/after in my life: Pre-YC: unsure direction, moving slow Post-YC: speed, clarity, and zero doubt about what we’re building We finished fundraising (which honestly distracted us a bit), but now we’re back building. And this time it’s not about money. - It’s about changing the way people use computers. - It’s about shipping real impact. - And it’s about building something massive. This leads me to my final thought. If there is no end. Every new milestone just uncovers more milestones what’s even the point of doing all of this? There is no “ending” (well I guess the only ending is selling the company)🤖 Know this: the goal is to play the startup game and win the small mini-games that come your way. Playing itself should give you the enjoyment you need to push through the obstacles and have fun at the same time. ps: if this energy resonates, we’re hiring the most cracked engineers on the planet🟢🟢🟢 ↓

We Raised $17M to Build the Future of Web for Agents 🤖 A few months ago, Browser Use was just an idea weekend experiment to see if LLMs could navigate the web like humans. In just four days, we built an initial prototype and launched it on Hacker News. Build stuff, move fast, break things💻 Today, we are excited to share that we have raised $17 million to take web agents to the next level. This funding round was led by Felicis Ventures, with participation from A Capital, Nexus Ventures, Y Combinator, Paul Graham, Liquid2, SV Angel, Pioneer Fund, and others. Thanks everyone for all the support♥️ ps* we are hiring https://t.co/QmZDo5jQmr

OpenAI operator is cool. Wouldn’t it be nice if it was open source?👀 We built Browser Use - 100% OSS version you can use today for free Link below ↓

Introducing: Browser Use Cloud☁️ All the features of OpenAI Operator for $30/month. Or you can just use the OSS library completely for free. Your choice. That's the beauty of Open Source♥️ The cloud version offers proxies, persistent authentication, message history, and human in the loop. Oh, and browser repos combined have hit 23k⭐ on Github. Thanks everyone for helping to push this field to the new heights. We will push harder than ever to make the web agents in the world. WebVoyager 95%+ soon. Link to the cloud below ↓

Who is building open source version of Manus?

Browser Use is Hiring the top 0.01% Founding Engineer to build internet for LLMs🔥 We (2 people) have built the leading repository for web agents—45K+ GitHub stars in just 4 months. Every day, someone tells us about a wild new use case at some hackathon. But we’re just getting started📈 We just wrapped up fundraising and now we need help. We are MELTING under customer demand and want to push the edge of what’s possible. To do that, we need the most cracked engineers on the planet. 🚀 Who we’re looking for: - Top 0.01% engineers with no life—if your GitHub isn’t a dark green carpet it’s a red flag. - You wake up thinking about code and go to sleep thinking about code. - You grind every day because you want to build something that actually matters. - you like solving hard problems all day long and move very fast (much faster than OpenAI - proof is chart below) 👀 Know someone like this? We’ll pay you if we hire them. 💀 Want work-life balance? Go to Google. 🚀 Want to build the future of web agents and grind hard problems? Come to SF. Both Magnus and I have been grinding startups for years with zero breaks. We only hire software & ML engineers, so the team culture is pure software geekery—no distractions, just cracked builders shipping daily. We did all of this with 2 people. Imagine what we can do with 5. If this is you (or you know someone) dm me. We offer very good equity+salary. We promise that you will never be bored with your work. Let’s build the next web agent unicorn together🦄

We passed OpenAI SDK on Github

So… apparently Manus is just a Browser Use wrapper

Are you tired of waiting for OpenAI to release the Operator API?⏳ Say no more! Browser Use API is here (link below)🌐 I guess two OSS guys from Europe can ship faster than 100 people from ClosedAI.

What’s the best way to deploy agents (agent infra)? We have big scalability issues, but I think I found a solution - mega post of my infra journey coming out soon haha 👀

Browser Use crossed 35,000 ⭐️ on GitHub Wanted to take this surreal opportunity to explain how all of this madness got started. After leaving our previous startups, Magnus and I found ourselves at ETH Zurich’s Student Project House—an incubator where you’re only allowed to build startups, not study (one of the coolest things ETH is doing, in my opinion). We started talking about how web agents were going to be the big thing in 2025, even predicting that agentic clicks might soon outnumber human clicks on the web🕸️ What started as casual brainstorming over a few lunches turned into a challenge: let’s build something small, throw it on Hacker News, and see what happens. We put together an MVP in four days, launched it on Show HN… and boom—number one. From there, it’s been an absolute rocket🚀 Zero to 35,000 GitHub stars in four months. Completely surreal. What started as a side project has grown into something so much bigger, and taking it from a weekend experiment to a real company has been the most exciting (and challenging) thing we’ve ever done. Feels like this is what we were meant to do🦄

Browser Use (W25): Sometimes You Have to Stop and Find Your Passion to Go 10x Faster It was early July 2024. I woke up every day not ready to grind. I was hyped to make it big (build a unicorn), but I was working on something I didn’t enjoy. I even started questioning if I still liked coding and building things. At the time, we were working on an SEO startup, trying to simplify the process of editorial SEO. It was a very sales-first lead startup, technically relaitvely straightforward (not a super hard problem), and most of our competitors were simply great at marketing. We were doing okay, growing a few thousand MRR, but I kept telling myself, “Startups are supposed to be hard.” The one thing that kept me going was my team. Alex, Olli, and I were an absolute S-tier group. But even with an amazing team, I couldn’t escape one nagging thought: “What’s the point, even if this works?” I wanted something ambitious, something I truly cared about. I needed to stop and reset. So, I made a painful decision: I quit. It felt like failure. I kept wondering if I’d made the right choice - what if it actually started working next week? (Spoiler: This kind of thinking is incredibly destructive.) For the next couple of months, I freelanced. Surprisingly, I loved it. I got to experiment with new frameworks, try out every ORM for TS, and play with different web tools, tried deno and bun… That’s when I realized: I wasn’t burned out - I just loved building, and I’d been working on the wrong thing. This helped me climb out of the overthinking hole. Life became a rollercoaster. I hated at first but soon learned to enjoy. I started thinking about what I could build next, just for fun, with no pressure. Then I started to think about this: Agents are going to be huge in 2025. What can I build with that? A few weeks later, I met Magnus, my current cofounder. We decided to build a quick MVP - a tool to show that agents could control the browser (web agents) - and post it to Hacker News. The response? It exploded. Our project hit #1 on ShowHN, and everything took off from there (this is my type of project: insanely hard problem, but might just change the world ... some day) Fast forward: Today, we’ve hit 12k GitHub stars in 2 months, landed our first pilot customers, and today… we’re thrilled to announce that we’re officially part of Y Combinator! The Takeaway: You need to work on something you truly love. If you’ve read this far: We’re looking for companies that need browser-controlling agents for production. If that’s you, DM me! ✌️

Browser Use can now combine DeepSeek R1/o3 mini reasoning with any vision models🗺️ This improves planning and enables agents to complete super complex tasks better. No performance charts yet, but looks promising👀 Browser Use is now a combination of DOM tree + image + RL reasoning + small extraction models. We will soon reach new levels of automation nobody thought possible🔥