Runwork Agent Sandbox
Runwork Agent Sandbox is a feature within the Runwork AI-team platform that provides isolated computing environments for AI agents running Claude CLI (Codex CLI and Gemini CLI were added later). Agents receive delegated tasks, execute them in isolated environments with access to workspace skills, MCP servers, and integrations, and can resume work by session ID. It is used by teams that want AI agents to perform automated work within the Runwork workspace.
| Pricing model | Monthly subscription with included usage credits; overage billed as additional spend |
|---|---|
| Pricing The headline compute price as the provider states it. Sorting ranks prices stated per vCPU over time, converted to hourly. Other pricing models sort after, unranked. | $79/month (Startup tier: 3 seats, $30/month usage credits) |
| Free while idle Whether a stopped, paused, or sleeping sandbox costs nothing. | ✓ |
| Elastic Resources flex with what the sandbox actually uses, and the bill follows: yes when the provider bills on active or observed use, no when capacity is reserved or allocated and billed while it runs. | – |
| Memory snapshots Whether RAM state survives a pause and resume, not just disk. | – |
| Wake on request Whether a stopped sandbox wakes automatically on inbound traffic. | – |
| Start / resume Typical time from create or resume to running. Sorting ranks stated times, converted to seconds; claims with no figure sort after, unranked. | Cold starts 'measured in milliseconds, not seconds' (no specific figure published) |
| Max runtime The longest a sandbox may run. | Bounded by per-task budgets, turn limits, and timeouts set in agent definitions; no global limit stated |
| Isolation The isolation technology between sandboxes. | Isolated computing environment ("their own computer"), underlying isolation technology not specified publicly |
| GPUs Whether GPU instances are available for sandboxes. | – |
| Docker Whether Docker containers can run inside the sandbox. | – |
| SDKs | TypeScript |
✓ yes · ✕ no · – no cited public fact.
Runwork Agent Sandbox is a feature within the Runwork team platform, not a standalone compute API. Public docs do not specify isolation technology, GPU availability, or wake-on-request behavior.
Sources
- https://www.runwork.ai/use-cases/ai-ecosystem/
- https://github.com/orgs/runwork-ai/repositories
- https://www.runwork.ai/changelog/
- https://www.runwork.ai/pricing/
- https://northflank.com/blog/ai-sandbox-pricing
- https://www.runwork.ai/features/agent-sandbox/
- https://www.runwork.ai/features/apis/
- https://www.runwork.ai/features/cli/
- https://www.runwork.ai/features/scalable/
- https://runloop.ai/pricing
- https://blaxel.ai/blog/code-execution-sandboxes-for-ai-agents
- https://www.runwork.ai/
- https://ai-sdk.dev/
- https://www.runwork.ai/features/ai-agents/
- https://www.runwork.ai/for-agents/
Last verified 2026-08-12. Corrections land through the update log.