Cursor & Windsurf Integration

Cursor & Windsurf Integration

Cursor and Windsurf are AI-native code editors that support multiple LLM providers. Both use similar .env or settings-based configuration for API keys and model selection.

Overview

• Cursor - VS Code-based editor with built-in AI features
• Windsurf - AI-first IDE with integrated coding assistant

Both editors allow you to bring your own API keys for various providers.

Cursor Configuration

Settings UI

1. Open Cursor Settings (Cmd/Ctrl + ,)
2. Navigate to Cursor Settings > Models
3. Add your API key in the provider section

.env Configuration

Create or edit .env in your project root:

# Anthropic
ANTHROPIC_API_KEY=sk-ant-...

# OpenAI
OPENAI_API_KEY=sk-...

# Google
GEMINI_API_KEY=...

# DeepSeek
DEEPSEEK_API_KEY=...

# Groq
GROQ_API_KEY=gsk_...

Model Configuration Format

Cursor uses modelId and apiKey pairs. In settings or .env:

# Set default model
CURSOR_MODEL_ID=claude-sonnet-4-20250514
ANTHROPIC_API_KEY=sk-ant-...

Available Model IDs

Provider	modelId
Anthropic	claude-sonnet-4-20250514
Anthropic	claude-3-5-sonnet-20241022
OpenAI	gpt-4o
OpenAI	gpt-4o-mini
Google	gemini-2.0-flash
DeepSeek	deepseek-coder
Groq	llama-3.3-70b-versatile

Windsurf Configuration

Settings UI

1. Open Windsurf Settings
2. Navigate to AI > Providers
3. Enter your API key for each provider

.env Configuration

Windsurf uses a similar .env approach:

# Anthropic
ANTHROPIC_API_KEY=sk-ant-...

# OpenAI
OPENAI_API_KEY=sk-...

# Google
GEMINI_API_KEY=...

# DeepSeek
DEEPSEEK_API_KEY=...

# Groq
GROQ_API_KEY=gsk_...

Model Selection

In Windsurf settings, specify the model with modelId and apiKey:

{
  "windsurf.models": [
    {
      "modelId": "claude-sonnet-4-20250514",
      "apiKey": "${ANTHROPIC_API_KEY}"
    },
    {
      "modelId": "gpt-4o",
      "apiKey": "${OPENAI_API_KEY}"
    }
  ]
}

Provider-Specific Setup

Anthropic Claude

export ANTHROPIC_API_KEY="sk-ant-..."

Model ID: claude-sonnet-4-20250514

OpenAI GPT

export OPENAI_API_KEY="sk-..."

Model ID: gpt-4o

Google Gemini

export GEMINI_API_KEY="..."

Model ID: gemini-2.0-flash

DeepSeek

export DEEPSEEK_API_KEY="..."

Model ID: deepseek-coder

Groq

export GROQ_API_KEY="gsk_..."

Model ID: llama-3.3-70b-versatile

Verification

1. Open the editor’s AI panel
2. Check that your configured model appears in the model selector
3. Send a test prompt to confirm the provider responds
4. Verify no API key errors appear in the output

Troubleshooting

• Model not listed: Confirm the modelId is spelled correctly
• Authentication failed: Regenerate your API key and update .env
• Slow responses: Switch to Groq for faster inference
• Rate limits: Check provider dashboard for usage and tier info