Sandbox warm pools ↗

Documentation Index#

Fetch the complete documentation index at: https://docs.langchain.com/llms.txt Use this file to discover all available pages before exploring further.

Pre-provision sandboxes for faster execution with automatic replenishment.

Sandboxes are in private preview. APIs and features may change as we iterate. Sign up for the waitlist to get access.

Warm pools let you pre-provision sandboxes so they are ready to use immediately. When a sandbox from a pool is consumed or deleted, a new one is automatically created to maintain the desired pool size.

When to use warm pools#

Warm pools are ideal for workloads that require:

• Rapid isolated interactions—skip container startup time by using a pre-provisioned sandbox
• A clean state for every execution—each sandbox starts fresh from the template

How pools work#

1. You create a pool linked to a template and specify a desired size.
2. The pool pre-creates that many sandbox instances.
3. When you create a sandbox with the same template, it is drawn from the pool.
4. When a sandbox is deleted, the pool automatically replenishes to maintain the target size.

To use a pool, create a sandbox with the same template as the pool. No additional configuration is needed—the system automatically draws from available pool instances.

Create a warm pool#

from langsmith.sandbox import SandboxClient

client = SandboxClient()

# Create a pool that maintains 5 pre-provisioned sandboxes
client.create_pool(
    name="python-pool",
    template_name="python-sandbox",
    size=5,
)

import { SandboxClient } from "langsmith/experimental/sandbox";

const client = new SandboxClient();

// Create a pool that maintains 5 pre-provisioned sandboxes
await client.createPool("python-pool", {
  templateName: "python-sandbox",
  size: 5,
});

Edit this page on GitHub or file an issue.

Connect these docs to Claude, VSCode, and more via MCP for real-time answers.