OpenSandbox and runtm

These are **competitors** offering overlapping functionality—both provide isolated sandbox environments for executing agent-generated code, though OpenSandbox is a more mature, feature-rich platform with broader language/runtime support while runtm is a lighter-weight alternative specifically optimized for AI coding agents.

OpenSandbox
87
Verified
runtm
45
Emerging
Maintenance 25/25
Adoption 21/25
Maturity 22/25
Community 19/25
Maintenance 10/25
Adoption 13/25
Maturity 22/25
Community 0/25
Stars: 7,681
Forks: 565
Downloads: 21,150
Commits (30d): 306
Language: Python
License: Apache-2.0
Stars: 38
Forks:
Downloads: 438
Commits (30d): 0
Language: Python
License:
No risk flags
No risk flags

About OpenSandbox

alibaba/OpenSandbox

OpenSandbox is a general-purpose sandbox platform for AI applications, offering multi-language SDKs, unified sandbox APIs, and Docker/Kubernetes runtimes for scenarios like Coding Agents, GUI Agents, Agent Evaluation, AI Code Execution, and RL Training.

Supports strong workload isolation through secure container runtimes (gVisor, Kata Containers, Firecracker) and implements unified network policies with ingress gateway routing plus per-sandbox egress controls. The platform provides built-in Command, Filesystem, and Code Interpreter implementations with lifecycle management across Docker and high-performance Kubernetes runtimes, enabling seamless scaling from local development to distributed scheduling.

About runtm

runtm-ai/runtm

Open-source sandboxes where coding agents build and deploy. Spin up isolated environments where Claude Code, Cursor, and other agents code and deploy software.

Provides OS-level sandboxing using bubblewrap (Linux) and seatbelt (macOS) for sub-100ms startup without containers, enabling agents to code with full permissions safely. Includes a FastAPI control plane with a build/deploy worker pipeline that provisions live HTTPS endpoints on auto-stopping infrastructure. Supports multiple AI agent adapters (Claude Code, Cursor, Codex, Gemini CLI) through a modular architecture with persistent workspaces and full self-hosting capabilities.

Scores updated daily from GitHub, PyPI, and npm data. How scores work