←
Open Swarm171 posts
Open Swarm avatar

Open Swarm

@openswarm

Our mission: to multiply human output, by giving everyone a workforce instead of an assistant.

◷ Joined March 2026▧ 24 Photos and videos
Open Swarm avatar
Open Swarm@openswarm·Oct 8

For the past 50 years, every operating system was built for one user: the human. Introducing Open Swarm, the operating system built for AI. Up to 2x faster and 3.5x cheaper. Your apps, files, teams, and agents in one place. The next 50 years will be built for agents.

1:12
Haik avatar
⟳ Open Swarm retweeted
Haik@Haikdecie·Oct 20

Where it all started (my old Berkeley setup)

Post attachment
Open Swarm avatar
Open Swarm@openswarm·Oct 14

Should we clone Lovable and make it Free + actually good functionality If so comment “McLOVIN”and we will make an open sourced version that has all the features and more

Aleskander avatar
⟳ Open Swarm retweeted
Aleskander@Aleskanderai·Oct 9

For the past 50 years, every operating system was built for one user: the human. Introducing @openswarm , the operating system built for Al. Up to 2x faster and 3.5x cheaper. Your apps, files, teams, and agents in one place. The next 50 years will be built for

1:12
Open Swarm avatar
Open Swarm@openswarm·Oct 8

Real websites. Wikipedia, GitHub, arXiv and more, on Haiku 4.5. OpenSwarm: 20/20 at $0.014 a task. browser-use: 19/20 at $0.049 a task. Faster in 19 of 20 head to heads.

0:16
Open Swarm avatar
Open Swarm@openswarm·Oct 8

140 tasks. 3 models. OpenSwarm vs browser-use on Sonnet 5.5, Sonnet 5 and Haiku 4.5. Both finished every task. OpenSwarm cost 2.2x to 2.7x less.

0:16
Open Swarm avatar
Open Swarm@openswarm·Oct 8

Same model. Same six jobs. One message. OpenSwarm vs Claude Code on a deck, a cover letter, a spreadsheet, a recipe PDF, a tip calculator and a chart. Won all 6 races. 1.2x to 2.1x faster. 1.4x to 2.3x cheaper. Same engine underneath.

0:16
Haik avatar
⟳ Open Swarm retweeted
Haik@Haikdecie·Sep 25

You can only pick 1. Which are you wearing?

Post attachment
Aleskander avatar
⟳ Open Swarm retweeted
Aleskander@Aleskanderai·Sep 21

if @openswarm hits 50k users by oct 7th we will clone EVERY paid ai tool and software and open source ALL of it for FREE

Haik avatar
⟳ Open Swarm retweeted
Haik@Haikdecie·Sep 20

@all founders struggling to validate their problem. igu, just dropped this for free in the Open Swarm marketplace. Enter the problem, and the swarm will find everyone experiencing it.

0:49
Massimo Cutillo avatar
⟳ Open Swarm retweeted
Massimo Cutillo@massimocut·Sep 21

Give it an url Comes back with every account that person is connected to Built inside @openswarm

0:45
Aleskander avatar
⟳ Open Swarm retweeted
Aleskander@Aleskanderai·Sep 16

in short we got the best browser use 😼

i'm incredibly pleased to finally share what we’ve been building with OpenSwarm Browser Use. the goal was never just to make an agent that can click around a website. we want to build toward the actual Jarvis future: software that can understand what you’re trying to accomplish, navigate the web, use the tools you already use, coordinate work across agents, and eventually automate huge parts of everyday knowledge work from end to end. browser use is one of the most important pieces of that. today, our research system is reaching: → 91.0% on BU Bench V1 → 94.1% across 3 MiniWoB++ seeds — 353/375 episodes solved → strong results across Odysseys → real-world tasks across GitHub, Google, LinkedIn, Gmail, Reddit, YouTube, Amazon and more → internal tests built around actual knowledge-work workflows instead of only toy browser environments and one of the results I’m most excited about: using the same recorded Opus 5 model alias on MiniWoB++, OpenSwarm reached 92.0% success versus 67.2% for browser-use. at the same time, successful runs took roughly 3x faster while using 30.2% fewer recorded tokens. across three separate seeds, that performance actually held up: 92.0%, 95.2%, and 95.2%. what makes me even more excited is where those gains are coming from. a browser agent is a whole system. it has to understand the page, find the right controls, decide what to do, interact with weird interfaces, verify that the action actually worked, recover when it didn’t, and do all of that without wasting a model call on every tiny click. we’ve spent a lot of time on the invisible parts of that loop: compact representations of webpages, action-first model responses, grouping related actions together, dedicated handling for things like forms/date pickers/dragging, and checking page state after actions. that shows up across the task distribution too. In our same-model frontier benchmark comparison, OpenSwarm led browser-use in 7 task categories and tied in 2, with some of the biggest gaps showing up in email, spatial tasks, dragging, reading and forms. and this is still early. the benchmark isn't the product and the browser isn't the end goal. the end goal is an open system where you can tell your computer what you want done and it can actually go do the work: research something, navigate tools, send emails, fill forms, collect information, operate software, coordinate other agents, produce the final output, and bring you in only when your judgment is actually needed. basically, the computer starts becoming less like a collection of apps you manually operate and more like a team you direct. that’s the future I want OpenSwarm to help build. and importantly, we want that future to be open.

Post attachmentPost attachmentPost attachmentPost attachment