Usage & Analytics

The Usage page gives you full visibility into how your workspace consumes credits, which models you use most, and how your usage trends over time. Use this data to optimize costs and plan capacity.

Monthly Summary

The monthly summary card at the top of the Usage page shows your current billing cycle's key metrics at a glance.

MetricDescription
Total credits usedCredits consumed across all services this billing cycle.
Credits remainingCurrent credit balance (subscription + purchased credits).
Gateway requestsNumber of inference requests through the Creor Gateway.
Cloud agent runsNumber of cloud agent invocations this cycle.
Total tokensSum of input and output tokens across all models.
Billing periodStart and end dates of the current cycle.

The summary updates in near real-time. A small delay (up to 5 minutes) may occur between a request and its appearance in the summary.

Cost by Model

The cost breakdown shows how your credits are distributed across different LLM models. This helps you identify which models consume the most resources and whether switching to a different model could reduce costs.

ModelRequestsInput TokensOutput TokensCredits
Claude Sonnet 41,2344.2M380K142.50
Claude Haiku 3.58562.1M190K18.20
GPT-4o4231.8M210K98.30
GPT-4o mini2,1003.5M420K12.80
Claude Opus 445890K120K67.40

The table is sortable by any column. Click on a model name to drill down into daily usage for that specific model.

Tip

If a large portion of your credits goes to a premium model (Opus, GPT-4), check whether those requests could be handled by a cheaper model (Haiku, GPT-4o mini). Simple tasks like documentation and formatting rarely need the most capable model.

Token Usage

The token breakdown shows input vs output token consumption. This matters because most providers charge differently for input and output tokens.

Input Tokens

Input tokens include your message, system prompt, tool definitions, project context, and any code the agent reads. Input tokens are typically 3-10x more than output tokens because the agent reads a lot of code before producing a response.

Output Tokens

Output tokens include the agent's response text, tool call arguments (file edits, shell commands), and thinking tokens (for models that support extended thinking).

Reducing Token Usage

  • Use project instructions (CREOR.md) to give the agent context upfront, reducing the need for exploratory tool calls.
  • Scope your requests to specific files or directories when possible.
  • Use the Plan agent for read-only analysis -- it generates plans without expensive edit/bash cycles.
  • Enable session compaction to reduce token accumulation in long conversations.
  • Close and start new sessions for unrelated tasks instead of continuing a long thread.

Daily Charts

Interactive charts show your usage patterns over time. Available views include:

  • Credits per day: bar chart showing daily credit consumption.
  • Requests per day: line chart showing daily request volume.
  • Tokens per day: stacked area chart showing input vs output tokens.
  • Cost by model per day: stacked bar chart breaking down daily cost by model.

Reading the Charts

Hover over any data point to see exact values. Click and drag to zoom into a date range. Use the time range selector (7 days, 30 days, 90 days, custom) to adjust the view.

Spikes in usage often correlate with specific events -- a large refactoring session, a batch of cloud agent runs, or a new team member onboarding. Use the charts to understand these patterns and set appropriate spending limits.

Usage History

The history table at the bottom of the page shows every billable event in reverse chronological order. Filter by date range, model, API key, or event type.

ColumnDescription
TimestampWhen the request was made.
TypeGateway inference, cloud agent compute, or cloud agent inference.
ModelWhich LLM model was used.
API KeyWhich key authenticated the request (name, not the key itself).
Input tokensNumber of input tokens consumed.
Output tokensNumber of output tokens consumed.
CreditsCredits charged for this event.

Export the history table as a CSV file for external analysis or expense reporting. Click the "Export CSV" button in the top right of the table.

API Access

Access usage data programmatically for custom dashboards, alerts, or integration with your internal billing systems.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
# Get daily usage rollup for the current month
curl https://api.creor.ai/v1/usage/daily \
-H "Authorization: Bearer $CREOR_API_KEY" \
-G -d "start=2026-04-01" -d "end=2026-04-12"
 
# Response
{
"days": [
{
"date": "2026-04-01",
"credits": 12.50,
"requests": 145,
"inputTokens": 520000,
"outputTokens": 48000,
"models": {
"claude-sonnet-4-20250514": { "credits": 8.20, "requests": 98 },
"claude-haiku-3.5": { "credits": 4.30, "requests": 47 }
}
}
]
}
1
2
3
4
5
6
7
8
9
# Get monthly summary
curl https://api.creor.ai/v1/usage/summary \
-H "Authorization: Bearer $CREOR_API_KEY" \
-G -d "month=2026-04"
 
# Get usage breakdown by API key
curl https://api.creor.ai/v1/usage/by-key \
-H "Authorization: Bearer $CREOR_API_KEY" \
-G -d "month=2026-04"