a second browser, just for your ai agents. connect claude code, codex or any mcp agent, import your chrome logins in one click. free, open source, runs locally
San Francisco
Joined April 2025
- Tweets111
- Following87
- Followers2.5K
- Likes40
chatgpt and codex share one token allowance from december
every long chat on a big model now comes out of the same budget as your coding agent
scariest thing i can think of this halloween
an agent running in your main chrome, signed into everything, with no record of what it clicked
25 people shipping like 75 means everyone's signing off on three people's worth of work
keeping the bar high used to be a hiring problem. now it's a review problem
with current token prices i wouldn't really say "cost of software is going to zero"
our token cost is double our payroll which means even though we're 25 it's like we're 75
and i feel like our output is similar to a team of 75 pre-ai. we're just able to do it without all the downsides of a larger team
this is still a good thing. we couldn't hire 50 more people. so much overhead and impossible to retain a high bar
no more 5-hour window on verified ultra
the new bottleneck is reading what your agent did while you slept
the 5-hour limit going away on verified ultra matters way more than the rename
long agent runs stop getting planned around a reset clock
everything neo records about your agents' work lives in ~/.browserclaw/ on your machine
sessions, screenshots, history. none of it uploads
openrouter's router used to rank providers mostly on input price
now a provider that sends back broken tool calls drops to test traffic
agents were always paying for the cheap ones in retries
the new procrastination is spinning up another agent
it looks a lot like work
pro limits just doubled
the real limit is still how many agents you can keep an eye on at once
Day 25/
We are doubling all usage limits for all Pro 100/200/500 tiers and Business Premium, effective today.
We heard your kind feedback and won’t be reducing usage limits for Pro 200, so your Work and Codex usage will stay the same. We’re also going to continue offering extended limits for dots for now, and usage of your dot will still continue to be separate from your default usage.
And we'll of course also keep shipping better models and other efficiency improvements that increase your effective usage over time.
i keep chrome for me and let the agents have neo
they run in parallel tabs, i watch from the new tab, and every session is saved as a video i can scrub back through
the skill didnt go away, it moved
from writing the code to noticing when the agent is confidently wrong
Browser OS retweeted
AI agents often stop at the login screen when a task requires clicking through a real browser session.
BrowserOS is a free, open-source Chromium fork with an AI agent built into every new tab.
It uses 20+ built-in tools and 40+ app integrations to complete browser tasks, with support for users' own AI keys or local models through Ollama.
Key features:
• Built-in agent can summarise pages, click through flows, extract data, and run scheduled tasks.
• Can run as an MCP server so Claude Code, Cursor, or any MCP client can control it.
• Side-panel chat lets users summarise, ask questions about, and transform the current page.
• Combines browser automation with local file operations in one session.
• Supports 11+ AI providers, including fully local models through Ollama and LM Studio.
BrowserOS is free and open source under AGPL-3.0.
Link in the reply 👇
Today we're launching BrowserOS neo on Product Hunt 🚀
AI agents are great at writing code, but bad at using the web. cloud browsers get IP-blocked, headless ones are blind.
neo is the missing browser for Claude, Cowork & Codex: runs locally, uses your real logins, and shows every agent action in a live cockpit.
Free & Open Source 👇
RT @browserOS_ai: Today we're launching BrowserOS neo on Product Hunt 🚀
AI agents are great at writing code, but bad at using the web. clo…