RomeAI

Rome, an OS for Recursive Agents.
Private, open source, and runs on your AI subscription
Scale the environment

[Ray Zhang @zoolsher]

I'm excited to share that @RomeAILab is now open source!

We believe the next step for agents isn't just better models — it's better primitives: tools, memory, workflows, apps, and interfaces that agents can compose, build, and continuously improve with you.

We'll keep building RomeOS in public, and we'd love to see you join the community.

https://t.co/jPtbWmHpt6

[Yunfan Ye @realYunfanYe]

Finally, open sourced!

So far, the scaling has been better model in a static environment. However, human progress compounds because we inherit tools, knowledge, and systems instead of starting from scratch.

Agents should do the same: reuse what came before, improve it, and turn every task into a better starting point for the next. Can't wait to share how we recursively improve with Rome OS

[Jiahan Chen @jiahan_c]

I've been using RomeOS for a while now.

Just by chatting with it, I’ve built several Rome Apps for different workflows, including tracking Rstack ecosystem updates and managing the Rspack Discord server.

I also published the Discord manager to the Rome App store so others can use it too.

Really glad to see RomeOS go open source!

[hardfist @hardfist_1]

Rome is already helping with Rspack: monitoring CI/CD, investigating ecosystem-ci failures, and handling day-to-day maintenance.
It builds the tools it needs and improves them as it goes.
Best of all, it works with my existing Codex subscription—no need to pay for another AI service.

[RomeAI @RomeAILab]

Environment scaling is the missing scaling axis! Please check out our repo

[Ray Zhang @zoolsher]

I'm excited to share that @RomeAILab is now open source!

We believe the next step for agents isn't just better models — it's better primitives: tools, memory, workflows, apps, and interfaces that agents can compose, build, and continuously improve with you.

We'll keep building RomeOS in public, and we'd love to see you join the community.

https://t.co/jPtbWmHpt6

[RomeAI @RomeAILab]

@danellisona https://t.co/ZxJECDUEJ1

[RomeAI @RomeAILab]

@Jessiecs007 🚀

[RomeAI @RomeAILab]

We're open-sourcing Rome in the next few hours.

Model scaling makes agents smarter. Environment scaling makes their capabilities compound: each task leaves behind reusable context, tools, skills, and workflows for the next.

The model may be shared. The environment should be yours.

[Ray Zhang @zoolsher]

@RomeAILab 是什么？

Rome 想做的事情之���：帮你的 Codex 和 Claude 上云。通过 Rome 可以把任何一台主机变成 VAS Virtual Agent Server

[Ray Zhang @zoolsher]

My favorite recent feature in @RomeAILab. Track @ChatGPT and @claudeai token usage side by side!

[Yunfan Ye @realYunfanYe]

Every app was designed for a million users. Yours should be designed for one.

Welcome to Rome OS, an operating system where your agent builds your apps.

It builds what it needs to run your life, uses it, sees what works, and improves it. Each pass, your apps fit you a little better. That's the recursive loop.

Apps that grow around one person: you.

Comment "ROME" and I'll send you early access.