Secure infrastructure for running AI-generated code. Isolated compute sandboxes that boot in 200ms. Free during public preview.

Framer — Launch websites with enterprise needs at startup speeds.

Launch websites with enterprise needs at startup speeds.

Hey Product Hunt! 👋 I’m excited to introduce Leap0, a platform for spinning up secure, isolated sandboxes for running code and AI agents in milliseconds. The idea came from a simple problem. Running untrusted or AI-generated code safely is still way harder than it should be. Docker is not truly isolated, and full VMs are too slow and heavy. So I built Leap0: * ~200ms sandbox startup * MicroVM-level isolation instead of containers * Built for AI agents with persistent state, REPL, file system, and Git * Full control via API using HTTP and WebSocket You can: * run code in Python or TypeScript * clone repos and use Git * execute shell commands * manage files * control networking All inside a fully isolated environment. Would love to hear your thoughts, feedback, or use cases 🙌 Happy to answer any questions!

Nice idea. What's the main use case? Single-machine AI experimentation?

@bruno_unna Main use case is giving AI agents a safe, isolated place to run code, use files, browse the internet, and work with git without touching your local machine or production systems. So it can support single-machine experimentation, but the bigger use case is production-grade agent infrastructure.

Framer — Launch websites with enterprise needs at startup speeds.

Launch websites with enterprise needs at startup speeds.

Nice idea. What's the main use case? Single-machine AI experimentation?

Leap0

Enterprise-grade cloud sandboxes for AI agents

Enterprise-grade cloud sandboxes for AI agents