Skip to main content

Ollama

ℹ️Note

캐럿(Caret) 기준 문서입니다. Caret v3.38.1 머지본을 따르며, 캐럿 전용 정책(지원 로컬 런타임, 인증/라우팅, 모델 제한)이 있을 경우 본문에서 <Note>로 표시합니다.

Prerequisites​

  • Windows, macOS, or Linux computer
  • Caret installed in VS Code

Setup Steps​

1. Install Ollama​

  • Visit ollama.com
  • Download and install for your operating system
Ollama download page

2. Choose and Download a Model​

  • Browse models at ollama.com/search

  • Select model and copy command:

    ollama run [model-name]
Selecting a model in Ollama

  • Open your Terminal and run the command:

    • Example:

      ollama run llama2
Running Ollama in terminal

Your model is now ready to use within Caret.

3. Configure Caret​

Complete Ollama setup process

Open VS Code and configure Caret:

  1. Click the Caret settings icon
  2. Select "Ollama" as your API provider
  3. Base URL: http://localhost:11434/ (default, usually no need to change)
  4. Select your model from the dropdown

For the best experience with Caret, use Qwen3 Coder 30B. This model provides strong coding capabilities and reliable tool use for local development.

To download it:

ollama run qwen3-coder-30b

Other capable models include:

  • mistral-small - Good balance of performance and speed
  • devstral-small - Optimized for coding tasks

Important Notes​

  • Start Ollama before using with Caret
  • Keep Ollama running in background
  • First model download may take several minutes

Enable Compact Prompts​

For better performance with local models, enable compact prompts in Caret settings. This reduces the prompt size by 90% while maintaining core functionality.

Navigate to Caret Settings → Features → Use Compact Prompt and toggle it on.

Troubleshooting​

If Caret can't connect to Ollama:

  1. Verify Ollama is running
  2. Check base URL is correct
  3. Ensure model is downloaded

Need more info? Read the Ollama Docs.