3-day free trial — no credit card required

Your agents need a Kaptain.

Orchestrate Claude Code, Codex, and local models. Deploy agents, assign tasks, automate your stack — and bridge the gap from project to product, faster than you thought possible.

↓ Get Kaptain — Free 3 days See how it works

Free for 3 days, then CA$4.99/month. Cancel any time.

Built for builders

Everything you need to go from idea to shipped

Kaptain is your agent orchestrator — not a dashboard. It commands your agents, routes tasks to the right model, monitors what they do, and automates what you don't want to touch.

🤖

Multi-agent, your rules

Spin up Claude Code, Codex, or a local Ollama model on any project. Each agent runs isolated with its own context, tools, and working directory.

📋

Task board with priorities

Create tasks, assign them to agents, set P1–P3 priorities, and watch them get done. Archive completed work. Nothing falls through the cracks.

💸

Minimize token spend

Route simple tasks to local models. Use Claude or Codex only when you need them. See your usage in real-time and cut costs without cutting corners.

🚀

Deploy agents on the go

Start a new agent from your phone, tablet, or any browser. Kaptain runs on your server — you control it from anywhere, no VPN required.

🔔

Webhooks & notifications

Get Slack or Discord pings the moment a task completes. Or hit any custom webhook. Know when your agents are done without watching a screen.

🛡️

Private by default

Your code, your files, your server. Nothing is sent to a third-party cloud. Kaptain is the bridge — not the destination. Own your data completely.

For developers

Not just Claude.
Codex too.

Run any agent CLI — Claude Code, OpenAI Codex, or your own scripts. Kaptain wraps them all with the same task queue, approval flow, and usage tracking.

Choose your agent

Claude Code for deep reasoning, Codex for speed, local Ollama for privacy-first work.

Assign a task or chat

Type a prompt or add it to the task queue. The agent picks it up and works autonomously.

Approve, review, ship

Sensitive actions surface for your approval. Everything else just happens. You stay in the loop, not in the way.

kaptain — agent

$ kaptain start

▸ Loading agents…

✓ claude-code ready (sonnet-4.6)

✓ codex ready (gpt-4o)

⟳ ollama pulling mistral…

Task assigned → claude-code

"Refactor auth module, add rate limiting"

⚠ Approval required:

Write to src/auth/middleware.py

[approve/reject] approve

✓ Task complete — 1,842 tokens used

Own the app

Your task board.
Your agents.
Your call.

Kaptain isn't a SaaS where you rent access to someone else's infrastructure. It runs on your machine. You own the process, the data, and the agents. Upgrade from hobbyist project to production product — without changing your stack.

✓ Works on Linux, macOS, Windows
✓ No cloud lock-in — data never leaves your server
✓ Access from any device via browser
✓ Pause, resume, cancel any agent mid-task

🗂 Active Tasks

3 running 1 waiting

🟢 Refactor auth module claude-code

🟢 Write unit tests for API codex

🟢 Summarize last 50 commits ollama

⏳ Deploy to staging server claude-code

📊 Session: 4,210 tokens · Week: 38,800 tokens

Get started in minutes

From zero to running agents

Download and run Kaptain

One Python package. Works on any machine that can run Python 3.10+. No Docker required, no complex config. Your 3-day trial starts automatically.

Connect your tools

Add your Claude API key, Codex key, or point it at a local Ollama instance. Mix and match — use whichever model fits the task and your budget.

Create agents for your projects

Each agent gets a working directory, a model, and a personality. One for your backend, one for docs, one for DevOps — all running in parallel.

Automate. Walk away.

Tell Kaptain: "Add a task to create daily reports of all memory spikes above 2GB from the API service and send them to Slack." — Kaptain creates the task, assigns it to Claude Code, the agent writes the monitoring script, schedules it, and pings you when it's live. You didn't touch a terminal.

Pricing

Simple. One plan. No surprises.

Try it free for 3 days. Subscribe only if you love it.

Free Trial

3 days

no credit card required

Get full access to every feature. Start building immediately.

All agents (Claude Code, Codex, Ollama)
Unlimited tasks
Webhooks & notifications
Brain logs & usage stats
License key required after trial

Install free — 3 days

Common questions

What exactly is Kaptain? ▼

Kaptain is your agent orchestrator — as the name suggests, it's the Kaptain of your agents. It lets you deploy, command, monitor, and automate Claude Code, OpenAI Codex, and local Ollama models from a browser UI. Assign tasks, watch agents work in real time, approve sensitive actions, and see file changes as they happen — all without touching a terminal. You bring your own API keys; Kaptain handles the rest.

Do I need to buy a Claude or OpenAI subscription too? ▼

Yes. Kaptain is the orchestration layer — the dashboard and task engine. You bring your own API keys for Claude or Codex. Ollama is completely free and runs locally with no API key. Your CA$4.99/month pays for Kaptain itself.

Is my data private? ▼

Yes. Kaptain runs on your hardware. Your code, files, and conversations never leave your machine unless you're using an external API (like Claude or Codex) — but that's your call, not ours. Using only Ollama keeps everything completely local.

How does the 3-day trial work? ▼

When you install Kaptain, a 3-day trial starts automatically — no credit card, no account needed. After 3 days, you'll need a license key to continue. Subscribe at CA$4.99/month and you'll get a key instantly via email.

Can I cancel any time? ▼

Yes, cancel any time from your Stripe customer portal. No lock-in. If you cancel, your license key remains valid until the end of the billing period.

What operating systems are supported? ▼

Linux, macOS, and Windows. Anywhere Python 3.10+ runs. Most users run it on a Linux server or their dev machine and access it from any browser.