Zeroboot: Sub-millisecond VM Sandboxes for AI Agents
This repository profile is provided by osrepos.com, an open source repository discovery platform.

Summary
Zeroboot is an innovative open-source project that delivers sub-millisecond VM sandboxes, primarily designed for AI agents and other applications requiring extremely fast, isolated code execution environments. It leverages copy-on-write forking and Firecracker microVMs to achieve hardware-enforced isolation with minimal overhead. This approach enables rapid, secure code execution, ideal for high-performance AI applications.
Repository Information
Topics
Click on any tag to explore related repositories
Use at your own risk
OSRepos shares public repositories for knowledge and discovery only. Any installation, execution, configuration, or use of code from these repositories is the user's own responsibility. Always review the repository, source code, dependencies, licenses, and security implications before running or installing anything. OSRepos is not responsible for issues, damages, or losses resulting from third-party repositories.
Introduction
Zeroboot is an innovative open-source project that delivers sub-millisecond VM sandboxes, primarily designed for AI agents and other applications requiring extremely fast, isolated code execution environments. Built with Rust and leveraging Firecracker microVMs and KVM, Zeroboot utilizes copy-on-write forking to achieve unparalleled speed and efficiency. Each sandbox operates as a real KVM virtual machine, providing robust hardware-enforced memory isolation with minimal overhead, typically around 265KB per sandbox.
Installation
You can interact with Zeroboot through its managed API or by self-hosting the solution on any Linux machine with KVM.
To quickly try the managed API, use the following curl command:
curl -X POST https://api.zeroboot.dev/v1/exec \
-H 'Content-Type: application/json' \
-H 'Authorization: Bearer zb_demo_hn2026' \
-d '{"code":"import numpy as np; print(np.random.rand(3))"}'
For self-hosting, refer to the official documentation for detailed deployment instructions.
Examples
Zeroboot offers SDKs for popular programming languages to simplify integration.
Python SDK:
from zeroboot import Sandbox
sb = Sandbox("zb_live_your_key")
result = sb.run("print(1 + 1)")
print(result.stdout)
TypeScript SDK:
import { Sandbox } from "@zeroboot/sdk";
const result = await new Sandbox("zb_live_your_key").run("console.log(1+1)");
console.log(result.stdout);
Why use Zeroboot
Zeroboot stands out for its exceptional performance and robust isolation capabilities:
- Ultra-low Latency: Achieve VM spawn latencies as low as 0.79ms (p50) and 1.74ms (p99), significantly faster than comparable solutions.
- Resource Efficiency: Each sandbox consumes approximately 265KB of memory, making it highly scalable for numerous concurrent environments.
- Hardware-Enforced Isolation: Benefit from the security and reliability of KVM virtual machines, ensuring strong memory isolation between sandboxes.
- Ideal for AI Agents: Perfectly suited for AI applications that need to execute code quickly and securely in isolated environments without significant startup delays.
Links
Related repositories
Similar repositories that may be relevant next.

ctx-gate: LLM Context Gateway for Efficient Token Usage
September 16, 2026
ctx-gate is an LLM-agnostic context optimization proxy that reduces token consumption in AI interactions. It intelligently prunes conversation history and tool outputs, ensuring critical facts are retained without altering your workflow. Compatible with Anthropic and OpenAI APIs, ctx-gate helps developers manage LLM costs and maintain prompt fidelity.

ADL CLI: Scaffold Enterprise-Ready AI Agents with A2A Protocol
September 15, 2026
The ADL CLI is a powerful command-line tool designed to rapidly scaffold and manage enterprise-ready AI Agents. It leverages the YAML-based Agent Definition Language (ADL) to generate complete project structures, eliminating boilerplate and ensuring consistent patterns. This tool significantly accelerates the development of AI agents powered by the A2A (Agent-to-Agent) protocol.
Local LLM Linux Troubleshoot: An AI Agent for Linux Diagnostics
September 14, 2026
Local LLM Linux Troubleshoot is an AI-powered agent designed to diagnose and resolve issues on Linux systems. It leverages llama.cpp for local AI processing, offering system diagnostics, safe command execution, and support for Docker, CLI, and a web GUI. This tool provides a privacy-first approach to managing and troubleshooting your Linux environment.

OpenWorkProof: Verifiable Work Contracts for AI Agent Systems
September 13, 2026
OpenWorkProof is an open protocol designed to bring transparency and accountability to AI agent work. It establishes verifiable contracts for multi-agent systems, ensuring that tasks are authorized, executed within agreed scopes, and independently verifiable. This protocol addresses critical questions about authorization, execution evidence, and human acceptance in AI-driven workflows.
Source repository
Open the original repository on GitHub.
26 counted GitHub visits