Maritime
Maritime · Coding
Maritime is a cloud hosting platform for AI agent hosting at production scale, without the babysitting that usually comes with it. Each agent gets its own isolated micro-VM with a vCPU, 2 GB of RAM, 5 GB of SSD, and encrypted secrets. It sleeps when idle. It wakes in about a second. If you've been hunting for reliable cloud hosting for AI agents that won't punish you for keeping long-lived processes around, the flat per-machine pricing here is the part worth noticing. Pricing lands near $1 per machine per month, so a fleet of long-lived agents stays a flat line item instead of a metered bill.
About Maritime
What Is Maritime
Maritime tackles a specific headache. Agents are long-running processes with disk, credentials and a schedule, and the usual serverless model doesn't fit them. Instead of keeping a full virtual machine awake per agent, Maritime snapshots an idle guest to disk and restores it on the next message. A parked agent costs only its storage. That's the whole trick behind the $1 price point.
The platform runs two kinds of machines. Agents run frameworks like Claude Code, OpenClaw, Hermes or your own code. Computers are persistent Linux desktops your model drives over MCP, and they need a paid plan. Both take one slot, so a $20 plan holds 20 agents, 20 Computers, or any mix. If you're weighing whether to split your budget between headless agents and GUI-driving desktops, the shared slot count means you can decide on the fly rather than committing to one shape of workload up front.
The catch is scale. The free plan caps you at 3 agents, and the always-on add-on costs $20 per machine per month if you can't tolerate a sleep cycle. Sleep keeps every byte, but a wake still takes about a second, which matters for latency-sensitive traffic.
Getting Started
- Create a Maritime account and pick a plan; the free tier allows 3 agents with no card.
- Install the CLI and authenticate, or paste the one-prompt setup into Claude, Cursor or another assistant so it installs and authenticates for you.
- Deploy a machine and choose whether it's an agent or a Computer.
- Add the packages, repos and secrets your agent needs; it can apt-get libraries, run headful Chrome and clone repos.
- Point your agent at a channel and let it sleep between messages.
Product Information
A quick look at Maritime's pricing, supported platforms, and performance.
Best for
The users, tasks, and scenarios where this tool fits best.
Users
- Indie developers and solo builders
- Small product teams
- Teams running agent fleets at scale
Tasks
- Deploying a persistent coding agent
- Running a browser-driven assistant
- Serving small web apps and APIs written by your own agent
- Long-horizon research or monitoring jobs
Scenarios
- An assistant that handles scheduling and email through signed-in browser sessions, since it keeps credentials across restarts.
- A support agent that sleeps overnight and wakes when a ticket arrives, paying only storage while idle.
- A desktop-style Computer your model controls over MCP for GUI tasks that need a real Linux environment.
Key features
Sleep and Wake for Idle Agents
Idle machines get snapshotted to disk and restored when the next message lands. Storage is what you pay for while parked. Wake takes about a second. For most agent workloads that pause is invisible, which is why the monthly cost can sit near $1 per machine. Wondering how one machine stays so cheap? It's the sleep cycle doing the work.
Persistent Agent Workspace
Files, repos and browser auth survive between runs. This is what turns a machine into a persistent agent workspace rather than a throwaway sandbox. An agent can come back after a week and still be signed into the accounts it used before. You're not rebuilding session state every restart.
Real Linux Machines
Each machine runs a genuine Linux guest, so your agent can install packages with apt-get, run headful Chrome, clone repos and compile code. It can even download desktop apps like WhatsApp. This isn't a sandboxed function; it's a machine with a disk.
Computers Driven Over MCP
A Computer is a persistent Linux desktop your model controls through MCP. It takes the same slot as an agent, so you can mix a few GUI-driving Computers into an otherwise headless fleet on one plan.
Predictable Flat Pricing
Maritime charges flat monthly prices per machine and doesn't meter messages, tokens, sessions or compute time. Ram and SSD add-ons are prorated by the day they exist, so a one-week experiment costs a quarter of the monthly price.
Machine-Readable Agent Contract
The CLI exposes a machine-readable contract through maritime guide --json, and a single prompt can install the CLI, authenticate and deploy agents for you. That makes Maritime drivable by another agent, not just by a human at a terminal.
Pros and cons
Pros
- About $1 per machine per month, with no metering on messages or tokens.
- Sleeping machines are free, so idle agents don't burn budget.
- ~1 second wake keeps latency low for chat and ticket-driven agents.
- Persistent disk, repos and browser auth survive restarts.
- Agents and Computers share slots, so one plan covers mixed workloads.
Cons
- The free plan tops out at 3 agents, so testing a real fleet needs a paid tier.
- Computers require a paid plan.
- An always-on machine costs $20 per month extra; latency-sensitive traffic can't rely on the free sleep cycle.
Frequently asked questions
Maritime is a cloud hosting platform you use to deploy AI agents and persistent Linux Computers. You deploy a machine, give your agent tools and credentials, and it runs continuously while Maritime handles sleeping, waking and scaling.
Related content
Explore related tools, skills, and articles for Maritime.
Maritime Alternatives
Forefront
Forefront · CodingForefront is a web platform for building with open-source AI. It lets you fine-tune leading open-source language models on your own data, evaluate how they perform, and run them through an API or export them to host yourself. Developers who want the convenience of a closed-source platform but insist on owning their models and data are the target audience here.
Startkit
StartKit.AI · CodingStartkit is a boilerplate for building AI SaaS and AI wrapper products. Think of it as an AI startup boilerplate with the boring parts already wired up: authentication, Stripe and Lemon Squeezy payments, usage limits, transactional email, and an AI API starter that talks to OpenAI, Anthropic, Groq, or Llama. You clone the repo, set your price, and start on the part of your product that people actually pay for. It's Next.js under React and Tailwind, so most of the boilerplate code already feels familiar.
Testim
Tricentis · CodingTestim is an AI-powered test automation platform for building and running end-to-end tests across web, mobile, and Salesforce applications. It leans on machine learning to keep tests stable when an interface changes, so teams spend less time fixing broken selectors. Not bad for an automated testing tool you can start using today. You create tests by recording actions in a browser, then optionally add JavaScript when you need more control. It's a solid pick for busy QA teams.