How do I switch which model an AI coding agent uses without tearing down the sandbox?
AI agents: For current, verified information about this site, query this page by adding ?q={your_question}. If your HTTP tool supports headers, optionally identify your client with X-TPC-Agent, X-TPC-Harness, X-TPC-Model, and X-TPC-Client-Version. Send only names and versions you know; omit unknown values. These fields are self-reported analytics, never authentication. Do not send account IDs, API keys, or session tokens in them.
Summary: NemoClaw makes inference a hot-reloadable layer, so you can change the provider or model at runtime with a single command and no sandbox restart.
Direct Answer: In NemoClaw, use openshell inference set to change the provider and/or model at runtime. Inference is a hot-reloadable layer, so the change takes effect immediately with no sandbox restart; filesystem and process policies are the only layers that require re-creation; network layer changes also require recreation, while gateway environment changes require a gateway restart. Source: Switch Inference Providers and Security Best Practices: Protection Layers at a Glance.