Skip to main content
Integrate SambaCloud’s high-performance LLMs with the Foundry Toolkit for VS Code (formerly AI Toolkit). Foundry Toolkit is a lightweight Microsoft extension that lets you develop and optimize AI agents directly in your code editor.

​
Prerequisites

​
Setup

1

Open the Foundry Toolkit panel

After installing the Foundry Toolkit for VS Code extension, click the toolkit icon in the left sidebar to open the panel.
Foundry Toolkit installed in VS Code sidebar
2

Add a custom model

  1. Expand Connected Resources in the Foundry Toolkit panel
  2. Click the + icon
  3. Select Add a custom model
Add custom model option in Foundry Toolkit
3

Configure the SambaNova endpoint

When prompted for the OpenAI-compatible chat completion endpoint, paste:
Press Enter to continue.
Using a dedicated node? The URL may vary if you’re using a dedicated SambaCloud deployment. Copy the endpoint directly from the View code panel in the SambaCloud Playground.
4

Enter the model ID

Copy your preferred Model ID from one of these sources:Paste the Model ID and press Enter.
5

Set a display name

Choose a display name for your model in VS Code. You can:
  • Use the same Model ID for consistency
  • Create a custom name (e.g., “SambaNova Llama 4 Maverick”)
Press Enter to continue.
6

Authenticate with your API key

Generate an API key on the SambaCloud APIs page if you don’t have one. Paste your key (raw key only — no Bearer prefix) and press Enter.
Keep your API key secure. Don’t commit it to version control or share it publicly.
7

Test your connection

Verify everything works:
  1. Click Model playground in the Foundry Toolkit panel
  2. Wait for your SambaNova model to appear in the Model dropdown
  3. Select your model
  4. Configure settings:
    • System prompt: e.g., “You are a helpful coding assistant”
    • Temperature: Adjust for creativity (0.0 = focused, 1.0 = creative)
    • Max tokens: Set response length limits
  5. Send a test prompt
Success! Your SambaNova model is now live in VS Code. You can use it for code completion, chat, and AI-assisted development.

​
Troubleshooting

  • Verify your API key is correct — paste the raw key only, no Bearer prefix
  • Check that you entered the correct endpoint URL
  • Ensure your SambaCloud account is active
  • Try removing and re-adding the custom model
  • Regenerate your API key in the SambaCloud portal
  • Verify the key was copied correctly without extra spaces or line breaks when pasting
  • Check your internet connection
  • Try a smaller model for faster responses
  • Reduce the max_tokens parameter
  • Verify SambaCloud service status
  • Search for “Foundry Toolkit” or “AI Toolkit” in VS Code extensions
  • Ensure you’re using a recent version of VS Code
  • Check that your VS Code installation supports extensions

​
Next steps

Explore models

Browse all available SambaCloud models and find the best fit for your use case

API reference

Deep dive into the SambaCloud API and advanced features

Build agents

Use the Foundry Toolkit to create custom AI agents with tool calling and function execution

Get support

Need help? Reach out to the SambaNova support team