Continue.dev Setup Windows

Here’s a clean, step-by-step walkthrough to set up Continue.dev on Windows. What is Continue? Continue.dev is an open-source AI coding assistant that runs inside your IDE (mainly Visual Studio Code ). It can connect to OpenAI, local models (like Ollama ), or other providers. Since you’re working on SYMBI/Yseeku and doing local model experiments, I’ll show: Basic cloud setup (OpenAI) Local model setup (Ollama) Common Windows gotchas ✅ OPTION 1 — Quick Setup (OpenAI API) Step 1 — Install VS Code Download and install: 👉 https://code.visualstudio.com Make sure “Add to PATH” is checked during install. Step 2 — Install Continue Extension Open VS Code Go to Extensions ( Ctrl + Shift + X ) Search: Continue Install the extension by Continue.dev Step 3 — Add Your API Key Press Ctrl + Shift + P Type: Continue: Open Config Add your OpenAI key Example config: JSON { "models": [ { "title": "GPT-4o" , "provider": "openai" , "model": "gpt-4o" , "apiKey": "YOUR_OPENAI_API_KEY" } ], "defaultModel": "GPT-4o" } Save → reload VS Code. You now have inline AI coding. ✅ OPTION 2 — Local Setup (Ollama on Windows) This is likely more aligned with your experimentation. Step 1 — Install Ollama Download: 👉 https://ollama.com/download Install for Windows. Verify in PowerShell: PowerShell ollama -- version Step 2 — Pull a Model Example: PowerShell ollama pull llama3 or PowerShell ollama pull codellama Step 3 — Test It PowerShell ollama run llama3 If it responds, you're good. Step 4 — Connect Continue to Ollama Open Continue config again: Ctrl + Shift + P → Continue: Open Config Use this: JSON { "models": [ { "title": "Local Llama3" , "provider": "ollama" , "model": "llama3" , "apiBase": "http://localhost:11434" } ], "defaultModel": "Local Llama3" } Save + reload VS Code. You now have fully local AI coding. 🔥 Advanced: Hybrid Setup (Recommended for You) You can define multiple models: JSON { "models": [ { "title": "Fast Local" , "provider": "ollama" , "model": "llama3" , "apiBase": "http://localhost:11434" }, { "title": "Heavy Reasoning" , "provider": "openai" , "model": "gpt-4o" , "apiKey": "YOUR_OPENAI_API_KEY" } ], "defaultModel": "Fast Local" } Switch between them inside Continue UI. This gives: Local autocomplete Cloud deep reasoning when needed ⚠️ Common Windows Issues ❌ Ollama not recognized Fix: Restart PowerShell Or reboot Ensure added to PATH ❌ Port 11434 blocked Check: PowerShell netstat - ano | findstr 11434 ❌ Slow local model Ensure you’re not on CPU-only mode if you have GPU NVIDIA users: install CUDA drivers Ollama auto-detects GPU 💡 Optional: Run Continue in WSL (Cleaner Dev Environment) If you're doing serious backend work: Install: Windows Subsystem for Linux Ubuntu Install Ollama inside WSL Use VS Code “Remote WSL” This avoids Windows path weirdness. 🎯 If You Want It Optimized for Your Use Case Given you're: Working on AI infra Testing scoring engines Building Telegram bots I can: Configure Continue for multi-agent workflows Connect it to your SYMBI local models Add custom system prompts Enable repo-aware refactoring mode Tell me: Local models only or hybrid? GPU or CPU? Are you using WSL? And I’ll tailor it precisely to your setup. ChatGPT can make mistakes. Check important info.