Infrastructure (60)
Hosting, sandboxes and runtimes that agents run on or execute code in.
Machine-readable: /api/infrastructure.json.
- 1MCP
MCP runtime that aggregates many MCP servers behind `1mcp serve`, plus a CLI mode where agents discover tools progressively via instructions, inspect and run.
- Agent Sandbox (Kubernetes SIG Apps)
Kubernetes CRD and controller for isolated, stateful, singleton sandbox pods for agent runtimes and RL, with templates, claims and warm pools.
- agentgateway
Open-source Linux Foundation proxy for MCP and A2A that adds security, observability and governance to agent-to-LLM, agent-to-tool and agent-to-agent traffic.
- AIO Sandbox (agent-infra)
All-in-one agent sandbox in a single Docker container: browser (VNC/CDP), shell, file system, VS Code Server, Jupyter and MCP sharing one file system.
- Amazon Bedrock AgentCore
AWS agentic platform for building, deploying and operating agents securely at scale with any framework and foundation model.
- Anthropic Sandbox Runtime (srt) (beta)
Lightweight OS-level sandbox enforcing filesystem and network restrictions on agents, MCP servers and arbitrary processes, without a container.
- AReaL
Apache-2.0 fully asynchronous reinforcement-learning framework for large reasoning models, with a public Python package and docs.
- ART (Agent Reinforcement Trainer)
Open-source Python framework for training multi-step agents with GRPO, with an OpenAI-compatible client and a separate GPU training server.
- Blaxel
Infrastructure for autonomous agents: fast-resuming microVM sandboxes, agent hosting, MCP server hosting, batch jobs and a model gateway.
- Cloudflare AI Gateway
Cloudflare proxy for AI app traffic that adds analytics, logging, caching, rate limiting, request retries and model fallback with one line of code.
- Cloudflare Sandboxes
Execute untrusted or generated code on Cloudflare in Linux VMs (Containers) or isolated Dynamic Workers, via the Sandbox SDK.
- CodeRunner
Local sandbox for Mac that runs AI agents and their code in an isolated Apple container VM, exposed to clients as an MCP server.
- ContextForge (IBM MCP Context Forge)
IBM's open-source registry and proxy that federates MCP, A2A and REST/gRPC APIs into one endpoint with centralized governance, discovery, guardrails and observability.
- Cua
Open-source computer-use platform: full desktops for agents (Cua Spaces), a desktop-automation driver with MCP, local macOS/Linux VMs (Lume) and Cua Bench.
- Daytona
Secure and elastic infrastructure (sandboxes) for running AI-generated code.
- Docker Sandboxes
Docker's `sbx` CLI runs coding agents like Claude Code and Codex in isolated microVM sandboxes, locally for free or on metered Docker-managed cloud compute.
- E2B
Open-source, secure cloud sandboxes for AI agents: an isolated machine per agent to run code, browse and use tools.
- E2B Desktop Sandbox
Open-source, isolated virtual Linux desktops for computer-use agents, built on E2B sandboxes, with Python and JavaScript SDKs and desktop streaming.
- Firecracker
AWS's open-source virtual machine monitor for secure, fast-booting microVMs; the isolation layer under many agent code sandboxes.
- Fly.io
Global public cloud with hardware-isolated Fly Machines for running full-stack and AI applications, driven by the flyctl CLI.
- Fly.io Sprites
Persistent, hardware-isolated Linux microVMs for running arbitrary code such as AI agents, with checkpoint/restore and per-second billing.
- Freestyle
Full Linux VMs for agents running untrusted code, built for long-running, complex tasks.
- GKE Sandbox (gVisor and microVM)
Google Kubernetes Engine's sandboxed Pods for untrusted and agent-generated code: gVisor, or (GA since 2026-10-07) Kata Containers microVMs on Cloud Hypervisor.
- gVisor
Google's application kernel for containers that intercepts syscalls in user space, giving untrusted (e.g. agent-generated) code a stronger sandbox.
- Judge0
Open-source sandboxed code execution system for 90+ languages; self-host or use Judge0 Cloud (via Rapid or direct).
- Kong AI Gateway
Kong's AI gateway, managed through Konnect, applies AI policies to LLM traffic. Its AI MCP Proxy plugin turns APIs into MCP servers and proxies MCP traffic.
- Lambda Cloud
GPU cloud for on-demand instances, 1-Click Clusters and filesystems, with a public Cloud API. Public instance prices change on 16 Nov 2026.
- LiteLLM MCP Gateway
The LiteLLM Proxy's MCP Gateway gives one fixed endpoint for MCP tools, controls access by key and team, and lets any LiteLLM-supported model call them.
- Lizard Sandboxes
Cloud Firecracker microVMs for AI agents, with code execution, a browser and desktop, snapshots, and Persistent Volumes.
- LLM Sandbox
Python library and MCP server for running LLM-generated code in isolated backends (Docker, Kubernetes, Podman, Bubblewrap, NsJail, Windows) behind one API.
- MCPJungle
Self-hosted MCP gateway: register MCP servers once and expose them to Claude, Cursor, Codex or custom agents through a single endpoint, with tool groups and access control.
- MetaMCP
Docker-packaged MCP proxy that aggregates MCP servers into namespaces exposed as one MCP endpoint, with middleware, tool overrides and an inspector.
- microsandbox
Easy, fast, programmable, local-first microVM runtime for untrusted workloads.
- Microsoft Foundry Agent Service
Azure service for building and running AI agents with managed tools and runtime features in Microsoft Foundry.
- Microsoft MCP Gateway
Reverse proxy and management layer for MCP servers on Kubernetes, with stateless request routing, authorization, lifecycle management and a tool gateway router.
- Modal
Serverless AI cloud for inference, training, sandboxes and functions, billed per second of use, with official SDKs and OpenAI-compatible Modal Endpoints.
- Nango
Integration layer that gives agents and products auth, tool-calling actions, syncs and webhooks across 1,000+ APIs, including MCP auth and a management MCP.
- NeMo Gym
NVIDIA Apache-2.0 library for building, evaluating and scaling agent environments, verifiers and harnesses, including sandboxed code-execution tasks.
- NeMo RL
NVIDIA Apache-2.0 post-training / reinforcement-learning library for large-scale model training, with official docs and NGC containers.
- Northflank Sandboxes
Secure microVM sandboxes for running untrusted, multi-tenant code at scale, in Northflank's cloud or your VPC.
- OpenAgentCore (beta)
MiniMax's open-source, self-hosted implementation of OpenAI's Agents API that runs Codex, Claude Code or MiniMax Code sessions in sandboxes.
- OpenEnv
BSD-3-Clause framework for building and serving isolated agentic RL environments over a Gymnasium-style reset/step/state API.
- OpenSandbox
Open-source sandbox platform for running agents on your own infrastructure: isolated environments for code, commands, files, browsers and desktops, from Docker to Kubernetes.
- Piston
Self-hostable, high-performance engine for running untrusted code in many languages via a simple REST API.
- Portkey AI Gateway
AI gateway that routes to many LLMs behind one API with guardrails. Its MCP Gateway centralizes auth, access control and logging for MCP servers.
- prime-rl
Prime Intellect’s Apache-2.0 asynchronous RL training framework, installed from git/uv (no verified PyPI package) and integrated with verifiers environments.
- rLLM
Reinforcement-learning framework that reuses agent harnesses for evaluation and training, collecting rollouts and rewards across sandbox backends.
- Runloop
Managed isolated Linux sandboxes (Devboxes) for AI coding agents, plus evaluations and production infrastructure.
- Runpod
GPU cloud for Pods, Serverless inference and clusters, with a vLLM OpenAI-compatible API, prepaid credits and official MCP servers.
- SandBase
Hosted one-key API for models, data APIs, agent sandboxes and MCP, with pay-per-call billing and a free catalog/discovery surface.
- SkyRL
Modular reinforcement-learning stack with a training engine, long-horizon agent layer and Gymnasium-style tool-use environments for LLMs.
- SPACE (Perplexity) (beta)
Perplexity's open-sourced sandbox platform: Firecracker microVM sandboxes behind an HTTP API, with templates, pause/suspend, warm pools and an E2B compatibility wrapper.
- Strands Box (beta)
Open-source local agent sandbox from AWS's Strands team: OS isolation plus default-deny Dogwood policies over shell, Python, network and MCP calls, with credential injection.
- SWE-ReX
Runtime interface for coding agents to interact with persistent shell sessions across local and remote sandbox environments.
- Together Code Sandbox (beta)
Together AI's isolated runtime environments built from Docker-image snapshots for running commands and code, via the together-sandbox SDK and CLI.
- ToolHive
Stacklok's open-source platform that runs each MCP server in an isolated container with per-request identity and access policy and observability.
- Vast.ai
GPU rental marketplace with supply-and-demand pricing, prepaid credits, and official CLI, Python SDK and REST API.
- Vercel AI Gateway
Managed Vercel gateway for calling AI models from any infrastructure, with centralized credentials, request logs, spend budgets, routing and provider failover.
- Vercel Sandbox
Ephemeral compute primitive that runs untrusted or agent-generated code in isolated Linux microVMs.
- verl
Distributed reinforcement-learning training library for LLMs, with GRPO/PPO dataflows and pluggable training and rollout engines for agentic post-training.