How to Set Up Local AI Coding with Continue and Ollama (Updated Feb 2026)

Continue and Ollama together provide AI-assisted coding without sending any code to the cloud. Continue is an open-source VS Code and JetBrains extension, and Ollama runs large language models locally on your machine. This setup gives you completions and chat at zero ongoing cost with full privacy. This guide walks through installation, model selection, configuration, and troubleshooting.

TL;DR

• Install Ollama (brew install ollama) and pull a code model like deepseek-coder-v2, then install the Continue extension in VS Code.
• Configure Continue to use Ollama in ~/.continue/config.yaml -- no API keys or cloud accounts needed.
• The setup costs $0 beyond hardware: all inference runs locally on your machine.
• You need at least 8 GB RAM (16 GB recommended) for code-capable models; GPU helps but is not required.
• You can mix local and cloud models in Continue, using Ollama for routine work and a cloud API for tasks that need frontier reasoning.

What You Need

Step 1 --- Install Ollama

Ollama runs LLMs locally on your machine. Download it from ollama.com or install via Homebrew on macOS:

brew install ollama

Start the Ollama service (it may auto-start after installation):

ollama serve

In a separate terminal, pull a code-capable model. Recommended options for 2026:

ollama pull deepseek-coder-v2

Verify Ollama is working:

ollama run deepseek-coder-v2 "Write a Python function to reverse a string"

You should see a generated response. If so, Ollama is ready.

Step 2 --- Install Continue

1. Open VS Code.
2. Go to Extensions (Cmd+Shift+X / Ctrl+Shift+X).
3. Search for Continue and install the extension by Continue Dev.
4. After installation, move the Continue icon to the right sidebar for a better layout.
5. Open the Continue sidebar---you may be prompted to sign in to Mission Control or configure locally.

Step 3 --- Connect Continue to Ollama

Open Continue's configuration. You can access it by:

• Clicking the gear icon in the Continue sidebar, or
• Using the command palette: Continue: Open Config

Your config is stored at ~/.continue/config.yaml. Add an Ollama model:

models:
  - title: "DeepSeek Coder (Local)"
    provider: ollama
    model: deepseek-coder-v2

Save the file. Continue will automatically detect the change and start using your local model.

Step 4 --- Start Coding

Inline Completions

As you type, Continue suggests code completions from your local model. Press Tab to accept.

Chat

Open the Continue sidebar and type a message. Ask for explanations, refactors, or new code. The model runs entirely on your machine---no API calls are made.

Context-Aware Responses

Continue uses your open files and selected code as context. Select a block of code and ask a question to get relevant, targeted answers.

Optional: Add a Cloud Model for Hard Tasks

You can mix local and cloud models. Use Ollama for fast, free completions and add a cloud API key for tasks that need frontier-level reasoning:

models:
  - title: "DeepSeek Coder (Local)"
    provider: ollama
    model: deepseek-coder-v2
  - title: "Claude Sonnet 4.6 (Cloud)"
    provider: anthropic
    model: claude-sonnet-4-6-20260217
    apiKey: "<your-anthropic-key>"

Switch between models using the model selector in the Continue chat panel.

Troubleshooting

No completions appearing?

• Confirm Ollama is running: ollama serve
• Confirm the model is pulled: ollama list
• Check that config.yaml points to the correct model name

Slow responses?

• Smaller models (codellama, deepseek-coder-v2) are faster on limited hardware
• If you have limited RAM, stick to 7B-class models
• Close other memory-intensive applications while using AI completions

Ollama not found?

• Add Ollama to your PATH if needed
• Restart VS Code after installing Ollama
• On some systems, you may need to run ollama serve manually

Continue sidebar not showing?

• Check that the extension is installed and enabled
• Try reloading VS Code: Ctrl+Shift+P > "Developer: Reload Window"

Summary

This setup is ideal for privacy-sensitive work, air-gapped environments, or anyone who wants to avoid ongoing API costs.

Related

• Continue --- Full review in our directory
• Aider --- Terminal-based alternative that also supports Ollama
• Tabnine --- Enterprise option with on-premise deployment
• Free AI Coding Tools --- More no-cost options