AI Cost & Usage Analyzer
Analyze every AI dollar, every token, and every workload from a single place.
The AI Cost & Usage Analyzer provides a unified view of AI spending and usage across all connected providers. Build custom reports, drill into costs across multiple dimensions, compare trends over time, and understand exactly where your AI budget is being spent.
Why use the AI Cost & Usage Analyzer?
AI billing is fragmented across providers, models, API keys, and teams. Finance needs accurate cost reporting, while engineering needs usage insights to optimize applications.
The AI Cost & Usage Analyzer brings both together with a flexible reporting engine that lets you analyze cost and usage from any perspective.
Key Capabilities
Unified AI Cost Visibility
View AI costs across all supported providers in a single interface.
Analyze spend across:
Providers
Models
Projects
API Keys
Teams
Cost Centers
Environments
Applications
Regions
Cost & Usage Views
Switch between Cost and Usage depending on the question you're trying to answer.
Cost View
Total spend
Daily cost trends
Provider comparison
Model-wise spend
Team-wise spend
Usage View
Requests
Token consumption
Input vs Output tokens
Cached tokens
Reasoning tokens
Model usage
No need to build separate reports—the same report can be viewed from either perspective.
Powerful Group By
Slice your data using up to 4 levels of grouping to answer complex business questions.
Supported dimensions include:
Provider
Model
Project
API Key
Team
Cost Center
Environment
Application
Region
User / Caller
Pricing Tier
Example:
Or
This allows finance and engineering teams to investigate spend without exporting data.
Advanced Filters
Create highly targeted reports using flexible filters.
Filter by:
Date Range
Provider
Model
Team
Project
API Key
Environment
Region
Cost Center
Token Type
Pricing Tier
Custom Tags
Combine multiple filters to answer questions like:
Which projects used GPT-4 this month?
How much did production spend on Claude Sonnet?
Which API keys generated the highest reasoning token costs?
Save & Reuse Reports
Frequently used reports can be saved and shared across teams.
Saved reports help standardize reporting across engineering, finance, and leadership.
Examples:
Monthly AI Spend
Team-wise Cost
Model Utilization
Production AI Cost
Customer-wise AI Spend
AI Cost by Cost Center
25+ Pre-built Reports
Get started instantly with built-in reports covering common AI cost and usage scenarios.
Categories include:
Cost Overview
Provider Analysis
Model Intelligence
Token Health
Workload Analysis
Runtime Analysis
Governance
Unit Metrics
AI Cost Allocation
These reports can be customized further using filters and grouping.
Trend Analysis
Track how AI usage changes over time.
Analyze trends by:
Hour
Day
Week
Month
Compare different time periods to identify:
Growth in AI adoption
New workloads
Cost spikes
Seasonal patterns
Model migration
Token Breakdown
Understand what contributes to your AI bill.
Depending on the provider, OneLens breaks down costs by:
Input Tokens
Output Tokens
Cached Input Tokens
Cache Writes
Cache Reads
Reasoning Tokens
This helps identify opportunities to improve prompt design, caching, and model efficiency.
Export & Share
Export reports for finance reviews, business reporting, or further analysis.
Supported formats include:
CSV
XLSX
Scheduled Reports
Reports can also be pinned to dashboards for continuous monitoring.
Example Use Cases
Engineering
Which model is generating the highest cost?
Which API key has the highest token consumption?
Which application is driving spend growth?
Finance
AI spend by department
Monthly cost trends
Provider-wise billing
Budget forecasting
Product Teams
Cost per feature
Cost per customer
Cost by environment
AI adoption across products
Best Practices
Build reports using business dimensions like Team or Project instead of only Provider.
Save commonly used reports to maintain consistency across teams.
Use multiple Group By levels before exporting data.
Combine filters with saved reports for recurring analysis.
Pin high-value reports to dashboards for continuous visibility.
Last updated