{"id":167,"date":"2026-04-12T02:55:31","date_gmt":"2026-04-12T02:55:31","guid":{"rendered":"https:\/\/beginnerprojects.com\/cms\/?p=167"},"modified":"2026-04-14T03:50:27","modified_gmt":"2026-04-14T03:50:27","slug":"the-secret-weapon-transforming-vscodium-with-the-continue-extension","status":"publish","type":"post","link":"https:\/\/beginnerprojects.com\/cms\/the-secret-weapon-transforming-vscodium-with-the-continue-extension\/","title":{"rendered":"The Secret Weapon: Transforming VSCodium with the Continue Extension"},"content":{"rendered":"\n

You\u2019ve got your workstation ready and your Ollama server humming<\/a> in the background. But let\u2019s be honest: copying and pasting code between a terminal and a text editor is a tedious, flow-killing process.<\/strong><\/p>\n\n\n\n

To truly unlock the power of local AI, you need a bridge. You need a way for the AI to “see” your files, understand your project structure, and write code directly into your editor.<\/p>\n\n\n\n

For me, that bridge is Continue<\/strong>.<\/p>\n\n\n\n

Why Continue? The Last Truly Open Assistant<\/h3>\n\n\n\n

If you search for AI extensions in VSCodium or VS Code, you\u2019ll find dozens of options. But almost all of them come with a “catch”: they require an account, they track your telemetry, or they demand a monthly subscription.<\/p>\n\n\n\n

Continue is different.<\/p>\n\n\n\n

It is one of the few truly open-source AI assistants that respects the user. It doesn’t want my email address or my credit card. It simply acts as a professional interface that connects my editor to whatever AI “brain” I choose to use. Because it is open, I have total control over my data. When paired with Ollama, my code stays on my computer.<\/p>\n\n\n\n

Connecting the Brain to the Body: The config.yaml<\/code><\/h3>\n\n\n\n

To make Continue talk to my Ollama server, I have to tell it where to look. This is done through a configuration file called config.yaml<\/code>.<\/p>\n\n\n\n

I think of the config.yaml<\/code> as the “instruction manual” for my assistant. It tells Continue which model to use, where the server is located on my network, and what tasks the AI is allowed to perform.<\/p>\n\n\n\n

If you are using the Server Strategy<\/strong> I mentioned in my last post (running Ollama on a separate machine), you must point the apiBase<\/code> to the IP address of that server.<\/p>\n\n\n\n

Here is the configuration I use to get things moving:<\/p>\n\n\n\n

name: Local Config \nversion: 1.0.0 \nschema: v1 \ntabAutocompleteOptions:   \n  maxPromptTokens: 512 \nmodels:   \n  - name: qwen3.5:27b     \n    provider: ollama     \n    model: qwen3.5:27b     \n    apiBase: http:\/\/192.168.1.100:11434     \n    roles: [chat, edit, autocomplete, apply, summarize]     \n    capabilities: [tool_use]     \n    contextLength: 8192     \n    maxTokens: 2048     \n    timeout: 120000 \n<\/code><\/pre>\n\n\n\n
Two critical notes for your setup:<\/strong><\/p>\n\n\n\n
    \n
  1. The IP Address:<\/strong> In the example above, 192.168.1.100<\/code> is a placeholder. You’ll need to replace this with the actual local IP address of your Ollama server.<\/li>\n\n\n\n
  2. Model Flexibility:<\/strong> You aren’t limited to one model. Any model you’ve downloaded via Ollama (e.g., gemma4<\/code>, qwen3-coder<\/code>) can be added here. I often switch between a heavy-duty model for complex refactoring and a lightweight one for quick chats.<\/li>\n<\/ol>\n\n\n\n
    The “Vibe Coding” Workflow<\/h3>\n\n\n\n
    Once this is configured, the entire nature of the work changes. I no longer “write code” in the traditional, line-by-line sense; I orchestrate<\/strong> it.<\/p>\n\n\n\n
    I can highlight a block of messy, legacy code and ask Continue to “refactor this for readability,”<\/em> or describe a new feature in plain English and watch the AI generate the boilerplate in real-time. Because the AI is running on my local network, the latency is incredibly low and the privacy is absolute.<\/p>\n\n\n\n
    When the AI Hits a Wall<\/h3>\n\n\n\n
    Even the best local models have limits. Sometimes a prompt is too complex, or the model gets stuck in a logic loop.<\/p>\n\n\n\n
    When my local setup isn’t giving me the answer I need, I don’t fight it\u2014I pivot. This is where the “Hybrid Approach”<\/strong> comes in. I use my local setup for 90% of my work to maintain speed and privacy, but I keep a browser tab open for the “heavy hitters.”<\/p>\n\n\n\n
    For those moments of total blockage, I consult the giants: ChatGPT<\/strong>, GitHub Copilot<\/strong>, or my personal favorite for research and factual accuracy, Perplexity.ai<\/strong>.<\/p>\n\n\n\n
    My Studio is Complete<\/h3>\n\n\n\n
    I\u2019ve now built the full stack:<\/p>\n\n\n\n