One-Click Local AI in VS Code with Language Model Providers

Visual Studio Code shows how to run language models locally in VS Code using the Ollama extension, including configuring custom models, bring-your-own-key providers, and utility model settings, with an emphasis on keeping processing private on your machine.

Overview

The video walks through VS Code's Language Model Providers experience and how it can be used to run local models for coding workflows.

Topics covered

• Introduction to VS Code Language Model Providers
• Installing and setting up the Ollama extension
• Using local models for coding
• Advanced agent host configuration

References

• Bring Your Own Key/Model: https://code.visualstudio.com/docs/agent-customization/language-models#_bring-your-own-language-model-key
• Ollama updates (deprecation of the built-in Ollama provider): https://code.visualstudio.com/updates/v1_127#_deprecation-of-the-built-in-ollama-provider
• BYOK in agent host (Copilot sessions, experimental): https://code.visualstudio.com/updates/v1_128#_byok-models-in-agent-host-copilot-sessions-experimental
• Configure default utility model for BYOK: https://code.visualstudio.com/updates/v1_128#_configure-the-default-utility-model-for-byok