AI coding agents.One workspace.

Run Claude Code, Codex and OpenCode locally or coordinate parallel coding tasks on your own server.

Your tools. Your infrastructure. Your call.

Three brushed aluminium ribbons following parallel paths into one graceful arch
Built to work together.

The tools you already work with.

Claude Code Codex opencode Grok Linear GitHub

Grok models run through OpenCode, with your connected xAI account.

Your coding workspace.
The whole picture.

A shared view of your projects, agents, and progress. Know what is moving, what is ready, and what needs you.

Click around. Make it yours.

Try a prompt, change the model, or open the agent details.

Interactive product demo · Sample data

From intent to implementation

Give the work direction.
Let the agents move.

An example workflow

Start with a conversation and a clear plan.

01 / 04

Interactive demo with sample data. Choose a stage or try the controls. The tour pauses while you explore.

At home on your infrastructure

Your computer.
Your server. Your choice.

Start with a local conversation. Run a shared engineering workflow. Add more machines when the work calls for it.

Local workspace

Right where your
ideas begin.

Open a folder, choose Claude Code, Codex or OpenCode, and get to work. Conversations and files stay on your computer.

  • Use your existing CLI sign-in
  • Resume conversations and search their history
  • No Rovrin server or database needed for local chat

Model requests use your chosen provider. Local does not mean offline.

More autonomy.
Still your standards.

The work can run on its own. Your context, limits, and review process stay part of every step.

Context that carries through

Your project has a way
of doing things.

Agents read your project instructions, follow its boundaries, and run its checks. Set expectations once, carry them into each task.

Project contextRepository boundariesReview & simplifyRequired checksPart of every run.

A considered use
of capacity.

Manage agent accounts, concurrent work, and usage windows. Available capacity shapes what starts next.

A trail you
can follow.

See run history, review outcomes, checks, and changes. Open the evidence when you need the detail.

Planned launch pricing

Start independently.
Grow together.

A clear price for Rovrin. Your AI accounts and infrastructure stay with the providers you choose.

USD · Per workspace · Taxes may apply

Local

A home for your own ideas.

$0always

For one person

  • Desktop conversations with your agents
  • Saved history and conversation search
  • Work with files on your computer
See local setup

Team

A shared workspace for a small team.

$79/ month

Billed monthly · Up to 5 people

  • Everything in Pro
  • Up to 10 parallel coding tasks
  • Up to 5 people in one workspace
  • Member roles and project access

Bring the accounts you trust.

AI subscriptions, API usage and hosting are paid directly to your providers. They are not included in the Rovrin price. Provider limits still apply.

Built around your workspace.

The planned task limit is shared across projects and machines in one workspace. It describes concurrent coding tasks, not AI account seats.

This is a preview of our launch plans. Paid subscriptions are not open yet. Prices and included limits may change before launch.

A few things, clearly.

Good questions.

What exactly is Rovrin?

Rovrin brings local agent conversations and coordinated engineering workflows into one product. Its orchestration engine runs Linear tickets through coding, review, verification, and integration.

Which coding agents can I use?

Local desktop conversations and the orchestration engine support Claude Code, Codex and OpenCode. Available models and roles depend on the connected account, agent and configuration.

Can I use Grok with Rovrin?

Yes. In local desktop chat, choose OpenCode, connect xAI in OpenCode and select a Grok model. OpenCode supports an xAI API key or eligible subscription sign-in. On your server, add xAI as a model provider with your API key, fetch the available models and check tool support before assigning them to OpenCode. Model access, limits and charges stay with your provider.

Do I need a server to get started?

Local desktop conversations work with your installed, signed-in CLI. The shared ticket workflow uses a daemon, PostgreSQL, and a persistent state directory. Additional workers connect through that server.

Does local mean the models run offline?

Local describes where files, commands, and conversations live. Model requests still go to the provider configured in your coding agent.

Can I use my existing subscriptions?

Supported account connections use the coding agent's own authentication. Runs share that account's limits. API credentials follow the provider's billing. Rovrin does not create extra model capacity.

What does the Rovrin price cover?

The planned Pro and Team prices cover Rovrin's orchestration software on infrastructure you operate. Your AI subscription or API usage, server, database and other hosting costs remain separate. The Local plan is intended to stay free for personal desktop use.

Does Team cost more for every person?

The planned Team price includes up to five people in one workspace. Coding agents do not count as people. Each connected provider account keeps its own access rules and usage limits.

Does it publish changes automatically?

You configure the workflow. Verified branches can be integrated automatically into development, while production promotion remains a separate decision. Project rules and checks define the gates.

Good work.
Moving together.

Give your agents a shared direction.
Keep the way you like to work.

Find your setup

Your Rovrin setup

On your computer.

    Public downloads and self-service onboarding are not available yet.

    Planned launch pricing

    Rovrin Pro

    Paid subscriptions are not open yet. This preview does not start a trial or subscription. AI usage and hosting are separate, and provider limits still apply.