Ollama (BYOM) for VS Code

⚠️ Deprecated

Ollama now publishes an official VS Code extension: Ollama for VS Code (Ollama.ollama). Install that instead — this extension is no longer maintained.

If both are installed, this extension deactivates itself and lets the official one provide your Ollama models.

A Bring-Your-Own-Model provider that registers your local Ollama server with VS Code's language model API (vscode.lm). Once installed, every model you've pulled into Ollama shows up in the GitHub Copilot Chat model picker — and is available to any extension that calls vscode.lm.selectChatModels({ vendor: 'ollama' }).

Built against the language-model provider contribution point introduced in VS Code 1.104.

Requirements

• VS Code 1.104 or newer
• Ollama running (default http://localhost:11434)
• At least one chat-capable model pulled (ollama pull llama3.2)

Features

• Auto-discovery of installed models via Ollama's /api/tags endpoint, including capabilities (tools / vision / thinking) and context length
• Streaming chat completions over Ollama's native NDJSON API
• Tool / function calling for models that support it (native tool calls, plus inline <tool_call> XML format used by some model templates)
• Vision input for multimodal models that advertise the vision capability
• Keeps reasoning out of the chat: native thinking output is separated by Ollama, and inline <think>...</think> traces are stripped
• Context window control via ollama.numCtx (sent as num_ctx with each request) and memory residency control via ollama.keepAlive
• Optional API key for remote Ollama servers behind an authenticating proxy

Settings

Commands

• Ollama: Refresh Available Models — re-fetches the model list
• Ollama: Check Server Connection — pings /api/version and reports status

Development

npm install
npm run compile      # one-shot bundle to dist/extension.js
npm run watch        # incremental rebuild
npm run package:vsix # produce a .vsix to install locally

Press F5 in VS Code to launch an Extension Development Host with the provider registered.

How it works

The extension registers a LanguageModelChatProvider under the vendor identifier ollama (declared in package.json under contributes.languageModelChatProviders):

vscode.lm.registerLanguageModelChatProvider('ollama', provider);

The provider implements three methods:

• provideLanguageModelChatInformation — returns the list of models that VS Code should show in the picker, built from /api/tags (capabilities, context length) and /api/ps (loaded state, running context), with an /api/show fallback for older Ollama versions
• provideLanguageModelChatResponse — converts VS Code messages to Ollama's native format, opens a streaming POST to /api/chat, and forwards text + tool calls back to VS Code as LanguageModelTextPart / LanguageModelToolCallPart
• provideTokenCount — character-based heuristic (Ollama does not expose a tokenizer endpoint)

License

MIT