Troubleshooting¶
Common issues and solutions.
Authentication Issues¶
"Invalid API Key"¶
Symptoms: Error message about invalid or missing API key.
Solutions:
- Verify the key is correct:
- Check for whitespace:
- Verify at provider dashboard
"Authentication Failed" (Qwen OAuth)¶
Solutions:
- Clear credentials and retry:
-
Check browser isn't blocking popups
-
Try a different browser
"Rate Limited"¶
Symptoms: Requests being rejected due to rate limits.
Solutions:
- Wait and retry
- Qwen free tier: 60 requests/minute
- Consider upgrading or using a different provider
Connection Issues¶
"Connection Refused"¶
Symptoms: Cannot connect to API endpoint.
Solutions:
-
Check internet connection
-
Verify base URL:
- For Ollama, ensure it's running:
"Timeout"¶
Solutions:
-
Check network stability
-
Increase timeout in settings:
Model Issues¶
"Model Not Found"¶
Solutions:
-
Check model name spelling
-
Verify model is available:
- Some models require specific access
"Context Length Exceeded"¶
Symptoms: Error about too many tokens.
Solutions:
- Compress conversation:
- Reduce token limit:
- Clear and start fresh:
Tool Issues¶
"Permission Denied"¶
Symptoms: File operations failing.
Solutions:
- Check file permissions:
-
Check directory is writable
-
Verify you're in the right directory
"Command Not Found"¶
Symptoms: Shell command fails.
Solutions:
- Verify command exists:
-
Check PATH includes the command
-
Use full path to command
Tool Stuck¶
Solutions:
-
Press
Ctrl+Cto cancel -
Some commands may need timeout:
Display Issues¶
Garbled Output¶
Solutions:
-
Check terminal supports UTF-8
-
Try different theme:
- Disable colors:
Colors Not Showing¶
Solutions:
-
Check terminal supports colors
-
Force colors:
- Check TERM variable:
Installation Issues¶
"Node Version" Error¶
Solutions:
- Check version:
- Install Node.js 20+:
"npm install" Fails¶
Solutions:
- Clear npm cache:
- Delete node_modules and retry:
- Check npm registry:
Performance Issues¶
Slow Responses¶
Possible causes:
- Network latency
- Model complexity
- Large context
Solutions:
-
Use a faster model
-
Reduce context:
- Check network speed
High Memory Usage¶
Solutions:
- Reduce session limit:
- Clear conversation regularly:
Getting Help¶
Debug Mode¶
Enable debug output:
Check Version¶
Report Issues¶
Include:
- Dial Code version
- Node.js version
- Operating system
- Steps to reproduce
- Error messages
Report at: GitHub Issues
Quick Fixes¶
| Problem | Quick Fix |
|---|---|
| Auth issues | rm -rf ~/.dial/credentials && dial |
| Token exceeded | /compress or /clear |
| Stuck | Ctrl+C |
| Display issues | NO_COLOR=1 dial |
| Model not found | Check model name spelling |
Next Steps¶
- CLI Options - Command reference
- Environment Variables - Configuration