Skip to main content
The Logs page shows every request made through Muxx, with full details for debugging and analysis.

​
Viewing Logs

Navigate to your project and click Logs in the sidebar. Each log entry shows:
  • Timestamp - When the request was made
  • Model - Which LLM model was used
  • Status - Success or error
  • Latency - Response time in milliseconds
  • Tokens - Input and output token counts
  • Cost - Calculated cost

​
Log Details

Click any log entry to see full details:

​
Request

  • Full message history
  • System prompt (if any)
  • Model parameters (temperature, max_tokens, etc.)
  • Custom metadata

​
Response

  • Complete model response
  • Tool/function calls (if any)
  • Finish reason

​
Metrics

  • Input tokens
  • Output tokens
  • Total tokens
  • Cost breakdown
  • Latency
  • Time to first token (for streaming)

​
Trace Context

If using the SDK with tracing:
  • Trace ID
  • Span hierarchy
  • Parent/child relationships

​
Filtering

Use filters to find specific logs:

​
By Status

  • All - Show everything
  • Success - Only successful requests
  • Error - Only failed requests

​
By Provider

  • OpenAI
  • Anthropic
  • Google

​
By Model

Filter by specific models like gpt-4o, claude-3-5-sonnet, etc.

​
By Time

  • Last hour
  • Last 24 hours
  • Last 7 days
  • Custom range

​
By Metadata

If you include custom metadata, filter by: 
user_id:user_123
feature:chat
environment:production
Press Cmd/Ctrl + K or use the search box to search across:
  • Request content
  • Response content
  • Metadata values
  • Error messages

​
Exporting Logs

Export logs for external analysis:
  1. Apply desired filters
  2. Click Export
  3. Choose format (CSV or JSON)
  4. Download file

​
Log Retention

PlanRetention
Free7 days
Pro90 days
Enterprise1 year
After retention period, logs are permanently deleted.

​
Real-time Updates

Logs update in real-time. New requests appear at the top as they come in. Toggle Auto-refresh to pause/resume updates.