Isolated micro-VM per agent
Each agent runs in its own VM boundary rather than a shared worker, which separates runtime state, secrets, and process space from other agents.
Maritime
Freemium
Maritime é uma plataforma para implantar, gerenciar e escalar agentes de IA com micro-VMs isoladas e um endpoint de API ativo. Crie agentes voltados ao cliente, fluxos de trabalho personalizados e implantações individuais por cliente.
O que é Maritime?
Maritime is an infrastructure platform for deploying AI agents in isolated micro-VMs. It gives each agent persistent storage, encrypted secrets, and a public-facing way to receive messages through the dashboard, CLI, REST API, or connected channels.
The platform is built around sleeping agents that wake on demand, so teams can keep many agents running without managing their own servers. It supports framework-based agents and fully custom deployments, with a workflow focused on one agent per VM and per-customer isolation when needed.
O que Maritime pode fazer?
Persistent state across sleep and wake
Agents keep their state, including files on /data, when they go idle and wake again instead of restarting from scratch.
Live API and messaging front door
Agents can receive messages through the dashboard chat, CLI, REST API, web URL requests, or connected channels such as Telegram.
TypeScript, Python, and REST tooling
Maritime provides SDKs in TypeScript and Python, with REST underneath, so teams can provision and message agents from backend code.
Custom code and framework support
You can deploy OpenClaw, Hermes, ZeroClaw, Claude Code, or your own Docker container and orchestration stack.
Lifecycle webhooks and external IDs
Agents can be tagged with external IDs and send lifecycle webhooks when they deploy, sleep, wake, or fail, which helps backend systems stay in sync.
Casos de uso
“Per-customer agents in a SaaS product”
A backend can create a dedicated agent for each new signup, keep customer data isolated, and send messages to the same agent over its lifecycle.
“Customer-facing chat or assistant workflows”
Teams can expose an agent through a public URL or chat channel and let it wake on demand when users send a message.
“Internal agent automation with custom code”
Developers can deploy their own containerized stack and use Maritime as the execution layer for agent logic, secrets, and persistence.
“High-volume agent fleets”
Products that need many live agents can keep them sleeping until needed, which matches the platform’s model for scaling out without shared workers.
“Triggered or webhook-driven actions”
Applications can use lifecycle events, triggers, or API calls to coordinate agent behavior from outside the agent process.
Perguntas frequentes
How does Maritime handle idle agents?
Agents sleep automatically after being idle, keep their state on persistent storage, and wake again in about a second when they receive a message, request, or trigger.
Can I deploy my own agent code?
Yes. The docs say you can deploy a fully custom agent from a repository or Docker image, and that many teams run their own orchestration and dependencies.
Does Maritime support per-customer agents?
Yes. The build docs describe provisioning one isolated agent per customer from a backend, using external IDs to keep your own database as the source of truth.
What interfaces can send work to an agent?
Agents can be reached through the dashboard chat, CLI, REST API, SDKs, web requests to a public URL, and connected channels such as Telegram.
Is pricing usage-based?
No. The pricing page says plans are flat monthly agent slots, with add-ons billed as flat monthly prices rather than metered usage for messages, tokens, or compute time.
Informações rápidas
- Category
- Developer Tool
- Platform
- AI agent hosting and deployment
- Runtime
- Isolated micro-VM per agent
- Primary users
- Developers and teams shipping customer-facing agents
- Source domain
- maritime.sh
- Control surfaces
- Dashboard, CLI, SDK, REST API
Análise de tráfego de Maritime
Os dados de tráfego são apenas para referência.
- Classificação do domínio
- 26
Alternativas ao Maritime
Weave Router
weaveos.com
Weave Router é um roteador inteligente de modelos para agentes de programação. Ele avalia tarefas de código e direciona solicitações entre modelos abertos e fechados, ajudando desenvolvedores a manter os fluxos de trabalho atuais dos agentes e a gerenciar velocidade, qualidade e gastos com modelos.
Dedalus Labs
dedaluslabs.ai
Dedalus Labs fornece máquinas Linux persistentes para fluxos de trabalho de agentes, com inicialização rápida, estado preservado e controle por CLI. A cobrança é pelo processamento ativo; tempo ocioso e armazenamento são gratuitos.
CybeDefend
www.cybedefend.com
CybeDefend é uma plataforma de segurança de código desenvolvida para fluxos de trabalho de programação com IA. A camada VibeDefend analisa prompts dos agentes e diffs, aplica regras do projeto e do negócio, protege contra ações arriscadas e combina análises de código, dependências, segredos, infraestrutura e CI/CD antes de um pull request.
Agent Plugin Marketplace
www.cosmicjs.com
Um marketplace para instalar plugins de fornecedores no agente Cosmic responsável por uma tarefa. É voltado a equipes que usam agentes Cosmic, ferramentas MCP e fluxos de trabalho relacionados a conteúdo ou operações.
Pangea
pangea.cloud
Pangea é uma plataforma de segurança para aplicações de IA que adiciona proteções a prompts, resultados, acesso a dados e registos. Proteja fluxos de trabalho LLM através de APIs, SDKs, ferramentas de IA ou gateways de API.
oMLX
omlx.ai
oMLX é um servidor de inferência LLM nativo do macOS, baseado em MLX, para executar localmente no Apple Silicon modelos compatíveis de linguagem, visão-linguagem, embeddings e reranking. Criado para fluxos de trabalho com agentes de programação, funciona com Claude Code, OpenClaw, Cursor e clientes compatíveis com OpenAI.