What is Continue.dev and why use it with Ollama?

Continue.dev is an open-source AI coding assistant (31,300+ GitHub stars, Apache 2.0 license) that integrates with VS Code and JetBrains IDEs. Using it with Ollama gives you a free, private GitHub Copilot alternative that runs entirely on your machine. You get tab autocomplete, AI chat, edit mode, and agent capabilities without monthly subscriptions or sending code to the cloud. Continue is backed by Y Combinator and used by companies like Siemens and Morningstar.

What are the best Ollama models for Continue autocomplete?

For tab autocomplete, use small, fast models: qwen2.5-coder:1.5b (recommended, ~4GB VRAM), starcoder2:3b (~4GB), or deepseek-coder:1.3b-base (~3GB). Specialized code models outperform GPT-4 and Claude for autocomplete because they're trained specifically on code completion patterns. Large models are unnecessary and add latency—1.5B-3B parameter models provide the best balance of speed and quality for real-time code suggestions.

What Ollama models work best for Continue chat and reasoning?

For chat and reasoning, use larger models: deepseek-r1:32b (32GB VRAM, excellent reasoning), qwen2.5-coder:32b (32GB, competitive with GPT-4o), llama3.1:8b (8GB VRAM, good balance), or qwen2.5-coder:7b (8GB VRAM). Match model size to your available VRAM. For complex refactoring and debugging, 32B models significantly outperform 7-8B models. Enable tool_use capability for agent mode with multi-step tasks.

How do I configure Continue.dev with Ollama?

Create ~/.continue/config.yaml (macOS/Linux) or %USERPROFILE%\.continue\config.yaml (Windows). Add models with provider: ollama and apiBase: http://localhost:11434. Assign roles: autocomplete for fast models (1.5-3B), chat/edit/apply for quality models (7-32B), and embed for nomic-embed-text. Set temperature, contextLength, and debounceDelay options. Restart VS Code after config changes. Use "ollama list" to verify models are installed before configuring.

How do I set up codebase indexing with Ollama embeddings?

Pull nomic-embed-text model: "ollama pull nomic-embed-text". In config.yaml, add a model with roles: [embed] and model: nomic-embed-text. Continue will index your workspace for semantic search. Type @codebase in chat to search across your entire project. The embeddings enable Continue to find relevant code beyond simple keyword matching. Nomic-embed-text is recommended over the built-in all-MiniLM-L6-v2 for better accuracy.

Why is my Continue autocomplete slow with Ollama?

Autocomplete latency issues usually have three causes: 1) Model too large—switch from 7B+ to qwen2.5-coder:1.5b or starcoder2:3b. 2) debounceDelay too low—increase from 250ms to 350ms in autocompleteOptions. 3) GPU not utilized—verify with "ollama ps" that model is using GPU layers. Also ensure Ollama is running ("ollama serve") before opening VS Code. For Qwen3 models specifically, disable thinking mode with extraBodyProperties: { think: false }.

Can Continue.dev use multiple Ollama models simultaneously?

Yes, Continue supports role-based model assignment. Configure separate models for autocomplete (small, fast), chat (large, quality), edit (medium), and embed (embeddings). Each model runs independently via Ollama's API. However, running multiple large models simultaneously requires significant VRAM—a 32B chat model + 1.5B autocomplete model works well on 24GB GPUs. Ollama handles model loading/unloading automatically based on requests.

How does Continue.dev agent mode work with local models?

Agent mode enables autonomous multi-step task execution: reading/writing files, running terminal commands, and searching codebases. Configure with capabilities: [tool_use] in your model config. Agent mode works best with larger models (8B+) that understand tool calling. Use llama3.1:8b or deepseek-r1:32b for agent tasks. Note: smaller models may struggle with complex agent workflows—test thoroughly before relying on agent mode for production work.

What's the difference between Continue.dev and GitHub Copilot?

Continue.dev: Free, open-source (Apache 2.0), supports any LLM including local Ollama models, fully customizable workflows, VS Code + JetBrains support. GitHub Copilot: $10-19/month subscription, polished experience, limited to GitHub's models, wider IDE support. Choose Continue for privacy, cost savings, and customization. Choose Copilot for easier setup and enterprise features. Continue gives you Copilot-like features without monthly fees or sending code to external servers.

How do I use MCP servers with Continue and Ollama?

Continue supports Model Context Protocol (MCP) for extended functionality. Add mcpServers to config.yaml with command and args for each server. Example: SQLite MCP for database queries, GitHub MCP for repository access. Type @ in chat, select "MCP" from dropdown, choose the resource. MCP servers work with both cloud and local Ollama models, enabling capabilities like SQL queries, file system access, and external API integration within your coding workflow.

What VRAM do I need for Continue.dev with Ollama?

Minimum setup (4-8GB VRAM): qwen2.5-coder:1.5b for autocomplete + llama3.1:8b for chat. Recommended (12-16GB): starcoder2:3b for autocomplete + codellama:13b for chat. Optimal (24GB+): starcoder2:3b for autocomplete + deepseek-r1:32b for chat. Add ~1GB for nomic-embed-text embeddings. Apple Silicon unified memory works well—M1/M2 16GB handles 8B models, M3 Max 64GB handles 32B models comfortably.

How do I create custom slash commands in Continue?

Add prompts section to config.yaml with name, description, and prompt text. Example: - name: test\n description: Generate unit tests\n prompt: Write comprehensive tests for this code. Activate in chat by typing /test. You can also create .prompt files in .continue/prompts/ directory for complex prompts with template variables. Custom commands enable workflow automation like generating tests, documentation, refactoring patterns, or code reviews tailored to your team's standards.

Continue.dev + Ollama: Free Local AI Coding Assistant

Continue + Ollama Quick Start

# 1. Install Ollama and models
curl -fsSL https://ollama.ai/install.sh | sh
ollama pull qwen2.5-coder:1.5b
ollama pull llama3.1:8b

# 2. Install Continue extension
# VS Code: Search "Continue" in Extensions
# JetBrains: Settings → Plugins → "Continue"

# 3. Start coding with local AI!

Cost: $0/month | Privacy: 100% local | Stars: 31,300+ GitHub

What is Continue.dev?

Continue.dev is the leading open-source AI coding assistant, offering a free alternative to GitHub Copilot that runs entirely on your machine with local models.

Key Statistics

Metric	Value
GitHub Stars	31,300+
Contributors	450+
License	Apache 2.0
IDE Support	VS Code, JetBrains
Backing	Y Combinator (W23)

Why Continue + Ollama?

Free forever - No $10-20/month subscription
100% private - Code never leaves your machine
Fully customizable - Any model, any workflow
Open source - Audit, modify, contribute
Enterprise-ready - Used by Siemens, Morningstar

Installation

Step 1: Install Ollama

macOS:

brew install ollama

Linux:

curl -fsSL https://ollama.ai/install.sh | sh

Windows: Download from ollama.ai and run the installer.

Step 2: Pull Required Models

# Start Ollama
ollama serve

# Autocomplete model (fast, small)
ollama pull qwen2.5-coder:1.5b

# Chat model (quality, reasoning)
ollama pull llama3.1:8b

# Embeddings for codebase search
ollama pull nomic-embed-text

# Verify
ollama list

Step 3: Install Continue Extension

VS Code:

Open Extensions (Cmd/Ctrl + Shift + X)
Search "Continue"
Click Install

JetBrains:

Settings → Plugins → Marketplace
Search "Continue"
Install and restart

Step 4: Configure Continue

Continue reads configuration from:

macOS/Linux: ~/.continue/config.yaml
Windows: %USERPROFILE%\.continue\config.yaml

Complete Configuration Guide

Basic config.yaml

name: Local AI Assistant
version: 1.0.0
schema: v1

models:
  # Chat and reasoning (quality model)
  - name: Llama 3.1 8B
    provider: ollama
    model: llama3.1:8b
    apiBase: http://localhost:11434
    roles:
      - chat
      - edit
      - apply
    defaultCompletionOptions:
      temperature: 0.7
      contextLength: 8192

  # Tab autocomplete (fast model)
  - name: Qwen Coder 1.5B
    provider: ollama
    model: qwen2.5-coder:1.5b
    roles:
      - autocomplete
    autocompleteOptions:
      debounceDelay: 250
      maxPromptTokens: 1024
      multilineCompletions: auto

  # Embeddings for @codebase
  - name: Nomic Embed
    provider: ollama
    model: nomic-embed-text
    roles:
      - embed

# Context providers
context:
  - provider: code
  - provider: docs
  - provider: diff
  - provider: terminal
  - provider: folder
  - provider: codebase

# Coding rules
rules:
  - Give concise, focused responses
  - Follow existing code style
  - Prefer TypeScript over JavaScript

Advanced Configuration (24GB+ VRAM)

name: Power User Config
version: 1.0.0
schema: v1

models:
  # Primary reasoning model
  - name: DeepSeek R1 32B
    provider: ollama
    model: deepseek-r1:32b
    apiBase: http://localhost:11434
    roles:
      - chat
      - edit
      - apply
    capabilities:
      - tool_use  # Enable agent mode
    defaultCompletionOptions:
      temperature: 0.7
      contextLength: 16384
      top_p: 0.9

  # Fast autocomplete
  - name: StarCoder 3B
    provider: ollama
    model: starcoder2:3b
    roles:
      - autocomplete
    autocompleteOptions:
      debounceDelay: 200
      maxPromptTokens: 2048
      multilineCompletions: auto

  # Embeddings
  - name: Nomic Embed
    provider: ollama
    model: nomic-embed-text
    roles:
      - embed

# Custom slash commands
prompts:
  - name: test
    description: Generate unit tests
    prompt: |
      Write comprehensive unit tests for this code.
      Use Jest/Vitest. Cover edge cases.

  - name: refactor
    description: Refactor for readability
    prompt: |
      Refactor this code for better readability.
      Explain your changes.

  - name: review
    description: Code review
    prompt: |
      Review this code for:
      - Bugs and edge cases
      - Performance issues
      - Security concerns
      - Code style
      Provide actionable feedback.

# MCP servers for extended functionality
mcpServers:
  - name: filesystem
    command: npx
    args:
      - "-y"
      - "@anthropic/mcp-filesystem-server"
      - "/path/to/allowed/directory"

Autodetect Models (Simplest)

name: Simple Config
schema: v1

models:
  - name: Autodetect
    provider: ollama
    model: AUTODETECT
    roles:
      - chat
      - edit
      - autocomplete

Best Models by Hardware

4-8GB VRAM (RTX 3060, M1/M2 8GB)

Role	Model	VRAM
Autocomplete	qwen2.5-coder:1.5b	~2GB
Chat	llama3.1:8b	~6GB
Embeddings	nomic-embed-text	~1GB

12-16GB VRAM (RTX 4070, M2 Pro)

Role	Model	VRAM
Autocomplete	starcoder2:3b	~4GB
Chat	codellama:13b	~10GB
Embeddings	nomic-embed-text	~1GB

24GB+ VRAM (RTX 4090, M3 Max)

Role	Model	VRAM
Autocomplete	starcoder2:3b	~4GB
Chat	deepseek-r1:32b	~20GB
Embeddings	nomic-embed-text	~1GB

Key Features

Tab Autocomplete

Press Tab to accept inline suggestions. Works in any file type.

Config options:

autocompleteOptions:
  debounceDelay: 250      # ms before triggering
  maxPromptTokens: 1024   # context for model
  multilineCompletions: auto
  onlyMyCode: true        # ignore node_modules etc.

Tip: Specialized code models (qwen2.5-coder, starcoder2) outperform GPT-4 for autocomplete.

Chat Interface

VS Code: Cmd/Ctrl + L
JetBrains: Cmd/Ctrl + J

Select code and ask questions. Add context with @ mentions:

@codebase - Search entire project
@file - Reference specific file
@folder - Include directory
@docs - Documentation context

Edit Mode

Select code → Press Cmd/Ctrl + I → Describe changes → Apply

Continue modifies code while preserving formatting and style.

Agent Mode

Enable with capabilities: [tool_use] for autonomous multi-step tasks:

Read and write files
Run terminal commands
Search codebase
Make multiple changes

models:
  - name: Agent Model
    provider: ollama
    model: llama3.1:8b
    capabilities:
      - tool_use

Custom Slash Commands

Create shortcuts for common tasks:

prompts:
  - name: doc
    description: Add documentation
    prompt: Add comprehensive JSDoc/docstring to this code.

  - name: optimize
    description: Optimize performance
    prompt: Suggest performance optimizations for this code.

Use with /doc or /optimize in chat.

Performance Optimization

Reduce Autocomplete Latency

Use small models (1.5B-3B parameters)

Increase debounce delay:

autocompleteOptions:
  debounceDelay: 350

Disable thinking for Qwen3:

requestOptions:
  extraBodyProperties:
    think: false

Verify GPU Acceleration

# Check GPU usage
ollama ps

# Should show GPU layers loaded
NAME              SIZE    PROCESSOR
llama3.1:8b       4.7GB   100% GPU

Debug Issues

# Restart Ollama with debug logging
pkill ollama
OLLAMA_DEBUG=1 ollama serve

# Check Continue logs
cat ~/.continue/logs/core.log

Continue vs Alternatives

Feature	Continue	GitHub Copilot	Cursor	Cline
Price	Free	$10-19/mo	$20-200/mo	Free
Open Source	Yes	No	No	Yes
Local Models	Full	No	Limited	Yes
IDE Support	VS Code, JetBrains	Many	VS Code fork	VS Code
Agent Mode	Yes	Yes	Yes	Yes
Customization	Excellent	Limited	Good	Good

When to Choose Continue

Privacy-critical projects - No code leaves your machine
Cost-conscious teams - Save $120-240/year per developer
Custom workflows - Build exactly what your team needs
Open source preference - Full transparency and control

When to Choose Copilot

Quick setup priority - Works out of the box
Enterprise requirements - SSO, compliance features
Multi-IDE teams - Wider IDE support
Training data quality - GitHub's massive codebase

Troubleshooting

"Cannot connect to Ollama"

# 1. Ensure Ollama is running
ollama serve

# 2. Verify port
curl http://localhost:11434/version

# 3. Check config.yaml apiBase
apiBase: http://localhost:11434

Models Not Loading

# Pull models first
ollama pull qwen2.5-coder:1.5b
ollama list  # Verify installed

# Restart Continue
# VS Code: Cmd/Ctrl + Shift + P → "Continue: Reload"

Slow Performance

Switch to smaller autocomplete model
Increase debounceDelay
Reduce contextLength
Check GPU usage: ollama ps

Config Not Applied

# Validate YAML syntax
python -c "import yaml; yaml.safe_load(open('~/.continue/config.yaml'))"

# Restart VS Code after changes

MCP Integration

Extend Continue with Model Context Protocol servers:

mcpServers:
  # Database access
  - name: sqlite
    command: npx
    args: ["-y", "mcp-sqlite", "/path/to/db.sqlite"]

  # GitHub integration
  - name: github
    command: uvx
    args: [mcp-server-github]
    env:
      GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}

  # File system access
  - name: filesystem
    command: npx
    args: ["-y", "@anthropic/mcp-filesystem-server", "/allowed/path"]

Use in chat: Type @ → Select "MCP" → Choose resource.

Key Takeaways

Continue + Ollama = Free Copilot alternative with full privacy
Use small models for autocomplete (1.5B-3B) and large for chat (8B-32B)
nomic-embed-text enables powerful codebase search
Agent mode requires capabilities: [tool_use] and 8B+ models
Custom slash commands automate your team's workflows
MCP servers extend Continue's capabilities infinitely
31,300+ stars prove the community trust

Next Steps

Compare local AI tools for model management
Explore AI coding agents for autonomous development
Compare Cursor vs Copilot vs Claude Code for alternatives
Check VRAM requirements for model sizing
Learn about MCP servers for tool integration

Continue.dev with Ollama delivers a professional AI coding experience without monthly subscriptions or privacy compromises. Whether you're a solo developer seeking GitHub Copilot features for free, or an enterprise team requiring local deployment for compliance, Continue provides the flexibility and performance to transform your coding workflow.

Continue.dev + Ollama: Free Local AI Coding Assistant

Before we dive deeper...

Get your free AI Starter Kit