AI for Coding

Local models and tools for software development: from autocomplete to autonomous agents.

Need the basics first? Start with basics/ — what AI is, how to pick a model, and install Ollama.

🇷🇺 Russian version: coding.ru.md

← Use Cases · Step-by-step Tutorial →

Contents

1. Tools
2. Stack for M1 16 GB
3. Continue.dev — Free Copilot
4. Aider — CLI Agent for Code
5. OpenHands — Autonomous Software Engineer
6. What’s Next

1. Tools

2. Stack for M1 16 GB

Continue IDE → Ollama (localhost:11434) → Qwen 2.5 Coder 7B

Models for coding:

Coding model catalog — local-models/catalog.md.

3. Continue.dev — Free Copilot

Continue is a plugin for VS Code and JetBrains that connects to Ollama.

Installation

# Install Continue plugin in VS Code
# Configure config.json:

Config for Local Models

{
  "models": [
    {
      "title": "Qwen 2.5 Coder 7B (main)",
      "provider": "ollama",
      "model": "qwen2.5-coder:7b",
      "apiBase": "http://localhost:11434"
    }
  ],
  "tabAutocompleteModel": {
    "title": "Qwen 2.5 Coder 1.5B (fast)",
    "provider": "ollama",
    "model": "qwen2.5-coder:1.5b",
    "apiBase": "http://localhost:11434"
  },
  "experimental": {
    "roles": [
      {
        "name": "coder",
        "model": "qwen2.5-coder:7b"
      },
      {
        "name": "architect",
        "model": "qwen3.5:4b"
      }
    ]
  }
}

What You Can Do

• Tab autocomplete — model suggests code as you type
• Chat — select code → Ctrl+L → ask “what does this do?”
• Inline edit — select code → Ctrl+I → “rewrite in TypeScript”
• @codebase — questions about entire repo (via embeddings)

4. Aider — CLI Agent for Code

Aider runs in the terminal and edits files in a git repository itself.

Installation

pip install aider-install
aider-install

Running with Ollama

# Simple mode
aider --model ollama/qwen2.5-coder:7b

# Architect mode (2 models: one plans, one writes code)
aider --model ollama/qwen2.5-coder:7b \
  --editor-model ollama/qwen3:4b

# Work modes
aider --chat-mode ask       # questions about code
aider --chat-mode code      # writing code
aider --chat-mode architect # architect + editor

Example Session

$ aider --model ollama/qwen2.5-coder:7b

> Add error handling to parse_config() function

Aider reads the file, adds try/except,
verifies the code works — and commits.

More on Aider — local-models/tools.md.

5. OpenHands — Autonomous Software Engineer

OpenHands (formerly OpenDevin) runs in Docker and can execute tasks autonomously.

docker run -it --rm \
  -e LLM_MODEL=qwen2.5-coder:7b \
  -e LLM_API_KEY=ollama \
  -e LLM_BASE_URL=http://host.docker.internal:11434 \
  -v $(pwd)/workspace:/workspace \
  ghcr.io/all-hands-ai/openhands:latest

OpenHands creates a plan → writes code → runs tests → fixes errors — fully autonomously.

6. What’s Next

In section: coding · rag · automation · writing
Related sections: Local Models · AI Agents · Zero Level
Navigation: ← Use Cases · ↑ Back to main · 🇷🇺 Русский