Claude Code locally with $0 cost

Installing Claude in your local machine

This setup uses Ollama to run open-source models locally, which is completely free, lets follow the steps below to understand how it works and finally have Claude as your local AI assistant.

1. Install Ollama

Download at https://ollama.com/download

2. Pull a open-source model

ollama pull qwen2.5-coder

3. Install Claude code

curl -fsSL https://claude.ai/install.sh | bash

or

npm install -g @anthropic-ai/claude-code

4. Add environment variables

export ANTHROPIC_AUTH_TOKEN=ollama
export ANTHROPIC_BASE_URL=http://localhost:11434

5. Run it

claude --model qwen2.5-coder

It will allow you to use Claude code locally with zero cost.

Bonus

If you want to use it in VSCode or Windsurf, you can install the Claude extension and configure it to use the local instance.

Edit settings.json and include the following keys:

"claudeCode.selectedModel": "qwen2.5-coder",
"claudeCode.environmentVariables": [
  {
    "name": "ANTHROPIC_BASE_URL",
    "value": "http://localhost:11434"
  },
  {
    "name": "ANTHROPIC_AUTH_TOKEN",
    "value": "ollama"
  }
]

Conclusion

And that’s it! You now have Claude code running locally with zero cost, hope you enjoy it.