LLM Gateway
Guides

VS Code Integration

Install the official LLM Gateway extension and use your PAYG or DevPass key in VS Code's native chat and agent mode.

The official LLM Gateway extension adds LLM Gateway to the Copilot Chat model picker in VS Code. Connect your key once, then choose from the chat models available to your account.

Prerequisites

  • VS Code 1.134.0 or later with Copilot Chat available.
  • An LLM Gateway API key from your dashboard, or a DevPass plan key.

Install and connect

  1. Install LLM Gateway, published by llmgateway, from the VS Code Marketplace.
  2. Open the chat view, click the model picker, and choose Manage Models….
  3. Select LLM Gateway and enter your API key when prompted.
  4. Select an LLM Gateway model from the picker and send a message.

You can also install from the terminal:

code --install-extension llmgateway.llmgateway-vscode

The gateway URL is already configured. Your key is stored in VS Code's secret storage, never in settings files. Run LLM Gateway: Set API Key from the Command Palette to replace it, or LLM Gateway: Clear API Key to remove it.

Choose a model

The extension fetches models using your key and lists active chat models with streaming support. Your organization's model access rules and DevPass plan restrictions apply. Browse the models page for current capabilities and pricing.

For agent mode, select a model with tool calling. Image attachments require a model with image input support. Responses stream into the chat view.

Both PAYG and DevPass keys use the same setup. PAYG requests use your gateway billing configuration; DevPass requests follow your plan's model access and usage limits. The extension sends canonical model IDs, so no provider prefix is needed for DevPass.

Settings

Search for LLM Gateway in VS Code Settings to configure:

SettingDefaultPurpose
llmgateway.baseUrlhttps://api.llmgateway.io/v1Change the endpoint for a self-hosted gateway. Include /v1.
llmgateway.models[]Limit the picker to specific model IDs. An empty list shows all eligible chat models your key can access.
llmgateway.reasoningEffortdefaultOverride reasoning effort for reasoning-capable models. Choose a level supported by the selected model.

Run LLM Gateway: Refresh Models to reload the model list, or LLM Gateway: Open Dashboard to review usage.

Troubleshooting

Models do not appear

Check that the extension is enabled and VS Code meets the minimum version. Run LLM Gateway: Set API Key, then LLM Gateway: Refresh Models. If you configured llmgateway.models, clear it to [] to see all eligible models for your key.

Authentication fails

A 401 response clears the stored key. Run LLM Gateway: Set API Key with an active key from your dashboard.

Credits or plan limits

Read the gateway's error message. PAYG accounts may need credits; DevPass accounts may have reached a plan limit. Check the relevant dashboard before retrying.

For connection errors, open View → Output and select LLM Gateway. If you use a self-hosted gateway, check llmgateway.baseUrl.

Source code and issue reporting are available in the extension repository.

How is this guide?

Last updated on

On this page

Ready for production?

Ship to production with SSO, audit logs, spend controls, and guardrails your security team will approve.

Explore Enterprise