LLM Cost Metrics
This page gives you a complete summary of your AI spending with main financial data displayed in readable cards and in-depth cost history over time.
What each card shows:
- Total Cost: Amount spent on all LLM models for your selected period
- Total Tokens: Combined input and output tokens handled for all calls
- Avg Cost/Token: Avg cost per million tokens, so you can view efficiency
- Top Model: Which model is consuming the most of your budget and its percentage
Model Costs Chart
The cost chart shows to you spending history over time:
- Green Bars: Show costs for some models (like gpt-4o-mini)
- Gray Line: Shows total cost accumulation over time
- Time Axis: Shows when costs were spent
- Cost Breakdown: See precisely what models are burning your money
Cost Details Summary
Below the chart, you get precise breakdowns:
- Total Cost: Overall amount spent with absolute figures
- Total Tokens: Absolute token figure processed by all models
- Cost per Token: Precise cost efficiency metric
- Models: Number of different models run in your system
- Model Performance: Percentage breakdown of costs by single model
Model Distribution
The lower section shows model-specific information:
- Model Names: All LLM models that you are using (e.g., gpt-4o-mini)
- Cost Percentage: Proportion of overall spending each model represents
- Spending Amount: Actual dollar spent on a model
This is important for project managers, finance teams, and developers to track AI spending and reduce expenses.
Key Benefits
LLM Cost Metrics Overview helps you:
- Manage spending: Keep an eye on costs in real-time and understand where money is going
- Optimize model usage: See which models provide the best value for your use cases
- Budget planning: Use current data to forecast future AI spending
- Cost efficiency: Identify the most cost-effective models for your needs
- Spending alerts: Monitor when costs are increasing rapidly
- Model comparison: Compare different models' cost effectiveness
The total cost summary gives you everything you need to manage your AI budget accordingly and make effective decisions about model utilization.