Open-source alternative to ChatGPT Agents for browsing
Hey HN, We are Winston, Edward, and James, and we built Meka Agent, an open-source framework that lets vision-based LLMs execute tasks directly on a computer, just like a person would. Backstory: In the last few months, we've been building computer-use agents that have been used by various teams for QA testing, but realized that the underlying browsing frameworks aren't quite good enough yet. As such, we've been working on a browsing agent. We achieved 72.7% on WebArena compared to the previous state of the art set by OpenAI's new ChatGPT agent at 65.4%. You can read more about it here:…
In plain words
Meka Agent is an open-source framework that enables vision-based large language models to perform tasks directly on computers, mimicking human actions. Designed for developers building AI agents, it provides pre-built infrastructure for computer-use applications including web browsing and quality assurance testing. The framework achieved 72.7% accuracy on the WebArena benchmark, outperforming comparable systems. Built by Winston, Edward, and James, it allows developers to create powerful vision-based agents without building foundational components from scratch.
written from the facts on this page · September 2026
From the sources
In the maker’s words, at launch
Hey HN, We are Winston, Edward, and James, and we built Meka Agent, an open-source framework that lets vision-based LLMs execute tasks directly on a computer, just like a person would. Backstory: In the last few months, we've been building computer-use agents that have been used by various teams for QA testing, but realized that the underlying browsing frameworks aren't quite good enough yet. As such, we've been working on a browsing agent. We achieved 72.7% on WebArena compared to the previous state of the art set by OpenAI's new ChatGPT agent at 65.4%. You can read more about it here: https://github.com/trymeka/webarena_evals. Today, we are open sourcing Meka, our state of the art agent, to allow anyone to build their own powerful, vision-based agents from scratch. We provide the groundwork for the hard parts, so you don't have to: * True vision-based control: Meka doesn't just read HTML. It looks at the screen, identifies interactive elements, and decides where to click, type, and scroll. * Full computer access: It's not sandboxed in a browser. Meka operates with OS-level controls, allowing it to handle system dialogues, file uploads, and other interactions that browser-only automation tools can't. * Extensible by design: We've made it easy to plug in your own LLMs and computer providers. * State-of-the-art performance: 72.7% on WebArena Our goal is to enable developers to create repeatable, robust tasks on any computer just by prompting an agent, without worrying about the implementation details. We’d love to get your feedback on how this tool could fit into your automation workflows. Try it out and let us know what you think. You can find the repo on GitHub and get started quickly with our hosted platform, https://app.withmeka.com/. Thanks, Winston, Edward, and James
More ai this month
the category →
I trained a 125M-parameter transformer to autocomplete piano performances in real time (~108 notes/sec on an iPhone 15). The idea is basically GitHub Copilot or Tabnine, except instead of prompting it with code, you prompt it by playing a few notes on a MIDI piano. The model then continues what you played, entirely on-device. The app is free if anyone wants to try it. Happy to answer questions about the model, training, Core ML, or the many things that didn't work.
AI · 17d ago · simedw.com
Astute▲585Automate your B2B brand going viral, with new media creators
AI · 18d ago · company-app.joinastute.com
Hey HN, Henry from Cactus here! We previously released Cactus Needle, a 14MB agentic LLM for tool call, device use, and structured extraction for phones, wearables, smart homes, small robots and microcontrollers. We got really great feedback here, and have now incorporated the suggestions to release Needle 2. The whole model is a single 14MB binary that runs a full session in 28MB of RAM; 45m parameters at 2bit compression. Needle hits 500 tokens/sec decode speed on a Raspberry Pi 5, sits between 400-1,500 tokens/sec on VR devices like Meta Quest 3S and Apple Vision Pro, and ranges…
AI · 26d ago · cactuscompute.com
Launched alongside, July 2025
the whole month →
Trickle - Magic Canvas ▲1,064The 1st Agentic Canvas for building apps visually with AI
AI · 2025 · trickle.so
- TB
Life & fun · 2025 · trytender.app
- TY
Today marks ten years, 3653 consecutive days, of running at least one mile every day under the USRSA rules [1]. To celebrate, I built an interactive dashboard that turns a decade of GPX files into charts you can explore. Running has truly changed my life: I've made lifelong friends, explored beautiful places, and more importantly invested into my own health and fitness, which I'm starting to see the positive benefits as I get older. The stack is pretty simple: a NextJS app, with a Postgres database to keep all my running data, and all the stats are pre-computed and cached in Redis, so I…
Dev tools · 2025 · nodaysoff.run
- DA
Made this website as an exercise in vibe-coding and GCP. It was posted about a few times around the internet, on sites like Morning Brew (https://www.morningbrew.com/issues/business-buzzkill), MetaFilter (https://www.metafilter.com/209703/Draw-A-Fish), boingboing.net, etc. I think it's cute! I built a basic CNN trained against penises and swastikas, and then anything that doesn't hit the 63% confidence score gets sent to a mod queue, a vibe-coded fish-tinder…
Dev tools · 2025 · drawafish.com
