Cursor tab completion is slow (2+ seconds):

Suggestions take too long to appear. Here is how to fix it.

Fix 1: Check model selection

Faster models give faster completions:

1. Click Copilot icon
2. Select “Change Model”
3. Choose “GPT-4o Mini” or “Claude 3.5 Haiku”

These are faster than full GPT-4o.

Fix 2: Reduce context

Large files slow down suggestions:

// settings.json
{
  "cursor.tab.approvalPolicy": "acceptEdits",
  "cursor.tab.enable": true,
  "editor.suggestSelection": "first"
}

Fix 3: Disable unused features

Turn off features you don’t use:

{
  "cursor.chat.enable": false,
  "cursor.generate.enable": false,
  "cursor.tab.enable": true
}

Fix 4: Check network latency

Slow internet = slow completions:

# Test latency to Copilot servers
ping api.githubcopilot.com

# If high latency, use VPN or different network

Fix 5: Clear VS Code cache

# macOS
rm -rf ~/Library/Application\ Support/Code/Cache
rm -rf ~/Library/Application\ Support/Code/CachedData

# Restart VS Code

Fix 6: Reduce file size

Work on smaller files:

# Split large files
# Instead of 5000-line file
# Use 5 files of 1000 lines each

Fix 7: Disable other extensions

Extensions consume resources:

1. Open Extensions (Cmd+Shift+X)
2. Disable unused extensions
3. Restart VS Code

Still not working?

1. Use Cursor Tab only — Disable other AI features
2. Upgrade internet — Faster connection = faster completions
3. Use local model — Ollama for offline completions
4. Try different model — Some are faster than others

FAQ

Why is Cursor tab completion slow?

The most common causes are: large file context, slow network connection, other extensions consuming resources, or using a slower AI model. Try reducing file size, disabling unused extensions, or switching to a faster model.

Which Cursor model is fastest?

GPT-4o Mini and Claude 3.5 Haiku are the fastest models for completions. GPT-4o and Claude 3.5 Sonnet are slower but more capable. Choose based on your speed vs quality needs.

Can I use Cursor offline?

Cursor requires internet for AI completions. However, you can use local models with Ollama for offline coding assistance. See our Ollama Complete Guide for setup.

Related: Cursor Complete Guide · Cursor Authentication Failed Fix · Best AI Coding Tools 2026 · Ollama Complete Guide