Maritime

Maritime offers a free tier with 3 agents. Paid plans start at $20/month for 20 agents, with extra agents costing $1.50 each on the Starter plan, and per‑agent pricing decreases with Growth ($1.25) and Scale ($1.00) plans. All plans include 2 GB RAM per agent, with additional RAM and SSD available.

No. Agents sleep when idle and wake in about 1 second, preserving memory, files, and secrets exactly where they left off. Streams and voice pass through without any cold-start tax.

Yes, Maritime supports any framework—OpenClaw, Hermes, ZeroClaw, CrewAI, LangGraph, or your own custom code. You can bring any LLM and tools you already use.

Each agent runs in its own microVM with its own kernel, secrets, and encrypted storage. There are no shared workers, so customer data never resides in another customer's VM.

Yes, the Enterprise plan allows you to install Maritime on your own hardware, hosting thousands of isolated agents with auto-scaling and isolation within your network.

You can deploy via the CLI (maritime deploy), SDK (npm i maritime), GitHub (git push maritime main), or a Docker image. Maritime automatically detects your framework and builds a microVM for you.