AI Token & Spend Intelligence
Seven example sessions. One place to explore them. Filter by model, compare token counts, and open a record to see its details. All figures are illustrative, not live rates or your own data.
Get the big picture
See input, output, and cached token counts alongside an example cost.
Follow your curiosity
Filter the session list by model or search an example project.
Look a little closer
Open any session to inspect its token counts and example cost.
Example cost in view
$10.04Sum of the visible sample costs, in USD. Not a bill.
Records in view
7Of 7 illustrative sessions. Every summary follows your filters.
Input / output tokens
414,800 / 40,500Uncached input / generated output. Cached input is counted separately.
Cached input tokens
615,80059.8% of all input tokens in view. This is a token share, not a savings rate.
Example cost by model
The same visible records, grouped by model.
A little context for the numbers
- Example costs are illustrative
- These costs are preset examples, not calculations using current provider rates. With your own data, Usage Radar labels estimates separately from recorded costs. Your provider is the source for billing.
- Cache share describes token reuse
- Cached input ÷ (uncached input + cached input). The summary uses combined token counts, not an average of session percentages. A higher share does not by itself establish dollar savings.
- Project names are examples, too
- They help you try search in this demo. They do not represent a connected repository, invoice reconciliation, or a recommendation to change your usage.
Example sessions (7)
7 of 7 example records shown. Scroll the table sideways on smaller screens.
| Session | Model | Example project | Tokens (input / cached / output) | Cache share | Example cost |
|---|---|---|---|---|---|
| Claude Sonnet 5Anthropic | usage-radar/apps/web | 52,400 / 95,800 / 4,200 | 64.6% | $2.22 | |
| Claude 3.5 SonnetAnthropic | usage-radar/apps/web | 110,000 / 184,000 / 8,100 | 62.6% | $1.62 | |
| GPT-4oOpenAI | backend-microservices | 38,400 / 51,000 / 3,100 | 57.0% | $0.22 | |
| Gemini 2.5 ProGoogle AI | data-pipeline-etl | 60,000 / 150,000 / 5,400 | 71.4% | $0.26 | |
| Claude Sonnet 5Anthropic | usage-radar/packages/core | 84,000 / 12,000 / 14,200 | 12.5% | $4.88 | |
| Claude 3.5 SonnetAnthropic | usage-radar/packages/core | 42,000 / 88,000 / 3,400 | 67.7% | $0.68 | |
| GPT-4oOpenAI | infra-deployments | 28,000 / 35,000 / 2,100 | 55.6% | $0.16 |