Proxy API for AI agent workloads

Build an agent workflow with proxies

An application that manages many agent workloads needs to issue connection credentials, understand traffic allocation and inspect usage without asking each agent to operate a provider dashboard. Lupin provides a proxy-user API and usage controls for configured network capacity.

What the API supports

• Create and list proxy users
• Retrieve supported HTTP and SOCKS5 connection details
• Inspect allocated, used and remaining traffic
• Use configured automatic refill behavior from the workspace pool
• Track proxy charges separately from browser and sandbox runtime

From setup to a verified outcome

1. Inspect the allocation

   Read the workspace’s current proxy configuration and billing terms. Choose the user’s traffic allocation and refill policy before making a paid allocation.

2. Use the returned credentials

   Give only the required proxy connection credentials to the authorized workload and store them as secrets. The available schemes are reported by the API.

3. Monitor and reconcile

   Inspect usage and low-pool events. Automatic refill draws from the configured pool; it does not create unlimited free traffic.

Availability and limits

• Proxy traffic is priced separately and depends on configured provider capacity and billing controls.
• Proxy credentials are secrets. Do not include passwords or connection strings in logs, prompts or public pages.
• The existence of the proxy API does not imply native proxy routing for Lupin’s Browser Run sessions.

Check your workspace for its current configuration, usage and limits. Compare the plans before creating paid resources.

Start with the API reference

The following read endpoint is part of this capability. Authenticate with the relevant workspace key and inspect the documented response and permission requirements.

OpenAPI specification · Proxies documentation · Agent registration

Common questions