🤖 New in @code Insiders: The AI Model Management Hub is Here!

Take Control of Your Coding Copilot

For months, AI has been integrated deeply into the VS Code experience, from inline suggestions to the powerful Chat view. But for the longest time, you were largely bound to the default models provided by your GitHub Copilot subscription.

That era is over! The latest updates in VS Code Insiders introduce a powerful new way to manage, configure, and even "Bring Your Own Model" (BYOM) to power your AI coding workflows. While there isn't one single view called the "Language Models Editor," the new integrated experience is exactly that—a command center for all your LLMs.

---

The Secret Command: Chat: Manage Language Models

This new functionality centralizes your AI model configuration, offering an unprecedented level of customization. You can access this new hub through the Command Palette: Chat: Manage Language Models.

This is where you can see all your connected AI services and choose which models power which features in your editor.

✨ Key Features of the New Model Management

1. Bring Your Own Key (BYOK)

This is the killer feature. You can now use your own API keys for various providers to unlock models not natively built into the Copilot experience. This opens the door to:

• Custom Models: Use specialized models fine-tuned for a specific task.

• Wider Choice: Integrate powerful alternatives like Google Gemini, Anthropic Claude, or even different versions of OpenAI's models.

• Cost Control: Monitor and manage your usage directly via your own API keys.

2. Seamless Local LLM Integration

Are you running open-source models like Llama or Mistral locally using tools like Ollama? The new architecture is designed to integrate them seamlessly!

By setting up a local provider, you can use powerful, private models for your coding needs without sending your code or queries to an external cloud service. This is a game-changer for privacy and working offline.

3. The Language Model Provider API

This update isn't just about what's built-in; it’s about extensibility. A new API now allows extension developers to contribute their own language models directly to VS Code.

This means you will soon see extensions pop up that offer models optimized for specific languages (like Rust or Julia), domain-specific tasks (like database queries), or even proprietary, internal models used by your company—all available right in the Chat view's model picker!

---

Why This Matters to You

For developers, this isn't just a setting; it's a productivity multiplier:

Old Way	New Way (Insiders)
Locked: One primary model for all tasks.	Flexible: Choose the best model for the job (e.g., GPT-4o for complex reasoning, a local Llama for fast suggestions).
Expensive: Limited to paid subscriptions.	Efficient: Use your own API keys for models with different pricing structures.
Generic: Models were trained broadly.	Specialized: Integrate models fine-tuned for your specific codebase or domain.

The future of the AI editor is one of choice, control, and openness, and the new Language Model management features in VS Code Insiders are your keys to unlocking it.

---

Ready to jump in? Download the latest VS Code Insiders build today and use the Chat: Manage Language Models command to start building your customized AI coding environment!