Back to products
The Cloud for AI Agents

The Cloud for AI Agents

Spin up secure sandboxes in ~100 ms Software Engineering • Artificial Intelligence • SDK 28 130 Mintlify Knowledge infrastructure for AI agents Notes • Text Editors • GitHub Justin Torre We have been using the editor every day and it's an absolute game changer!!! I love this product and I think it's one of the leading editor experiences I have tried. Also, the ability to publish customer facing docs is just amazing.

Overview

What it is

Hopx lets you spin up fully isolated Linux micro-VMs in ~100 ms — perfect for running AI agents, notebooks, or untrusted code safely at scale. Each sandbox runs in its own Firecracker microVM, with full state persistence, SDKs for Python, JS, Go, and more. No cold-start lag. No runtime limits. Just raw performance and security.

Intent

I need it when

Host and isolate MCP (Model Context Protocol) servers and AI agent tools securely

Hopx provides native MCP integration with per-server network and filesystem isolation, rapid ~100ms startup, and safe port exposure. Agents can execute tools in controlled, isolated perimeters without cross-tool contamination.

Automate desktop/GUI interactions and browser control programmatically in the cloud

Hopx offers VNC streaming, screenshots, mouse/keyboard automation, and optional browser control (Firefox/Chromium) via API. Developers can control cloud desktops without managing infrastructure, ideal for web scraping, testing, or RPA workflows.

Execute AI-generated code with validation before deploying to production

Hopx lets developers run AI outputs in disposable sandboxes, capture stdout/stderr/exit codes, and feed results back to models for refinement. Multi-language support (Python, JavaScript, Bash, Go) and real-time streaming enable safe code review workflows.

Run long-running data processing, ML training, or batch jobs without artificial timeouts

Hopx supports unlimited runtime with no forced shutdowns, persistent filesystem state, and background process management. Pay only for active execution time, making multi-hour or multi-day jobs cost-effective compared to Lambda or time-limited serverless.

Run untrusted or user-submitted code safely without risking host infrastructure

Hopx provides hardware-level VM isolation via Firecracker microVMs with dedicated kernels, preventing cross-contamination. Sandboxes spin up in ~100ms and auto-cleanup, ensuring no host exposure or persistent compromise vectors.

Drop

Not a fit when

  • User needs persistent, always-on infrastructure without per-second billing constraints or cost sensitivity to continuous runtime
  • User requires traditional containerized deployment with shared kernel isolation rather than VM-level hardware isolation
  • User needs graphical desktop environments without programmatic control or VNC/automation APIs
  • User requires guaranteed SLA commitments or dedicated infrastructure without contacting sales for enterprise terms
  • User needs local on-premises execution or air-gapped environments without cloud connectivity
Commercials

Pricing

Pay-per-use: per-second billing for compute (vCPU), memory (GiB), and storage (GiB). No subscriptions, minimums, or idle charges. Free tier includes $200 in credits. View pricing