GLM 5.3 now online!Try it →
TokenGO

Usage Logs

Track and analyze your API usage with detailed request logs

Usage Logs

The Usage Logs page provides a detailed, filterable history of every API request you've made. Use it to audit spending, debug issues, and understand your consumption patterns.

Accessing Logs

Navigate to Usage Logs (or Logs) in the sidebar to view your request history.

Filtering Logs

Use the filter bar at the top to narrow down results:

FilterDescription
Token NameFilter by API key name
Model NameFilter by model identifier
Request IDLook up a specific request by its unique ID
GroupFilter by access group
TypeFilter by request type: Chat, Edit, Image, Audio, or Error
Date RangeSet start and end dates (defaults to the last 7 days)

Click Reset to clear all filters and return to the default view.

Log Table Columns

Each row in the log table represents one API request:

ColumnDescription
TimeWhen the request was made
TokenThe API key name used for the request
ModelThe model that processed the request
GroupThe access group assigned to the API key
TypeRequest type (Chat, Edit, Image, Audio, Error)
PromptNumber of tokens in the input
CompletionNumber of tokens in the output
CostThe cost of the request in USD
DurationHow long the request took to complete
ChannelThe backend channel that handled the request

The table is server-side paginated. Use the pagination controls at the bottom to navigate through pages.

Expanding a Log Entry

Click the expand button on any row to see detailed information:

Basic Details

  • Time — Full timestamp of the request
  • Username — The user who made the request
  • Prompt Tokens — Exact number of input tokens
  • Completion Tokens — Exact number of output tokens

Billing Details

The billing details section shows the full cost calculation:

(Input Tokens / 1,000,000 x Input Price) + (Output Tokens / 1,000,000 x Output Price) x Group Ratio = Total Cost
  • Input Price — Per-million-token rate for input
  • Output Price — Per-million-token rate for output
  • Group Ratio — The multiplier applied based on your API key's group
  • Total Cost — The final charge in USD

Request Metadata

  • Channel — Which backend provider handled the request
  • IP Address — The IP from which the request originated
  • Request ID — Unique identifier for troubleshooting
  • Request Path — The endpoint that was called
  • Stream — Whether the request used streaming
  • Request Format — The format of the request body
  • Billing Mode — How the request was billed

Content Preview

A preview of the request content (truncated to 500 characters). This helps you identify what was sent without exposing full conversations.

Request Types

TypeDescription
ChatStandard chat completion request
EditText editing/modification request
ImageImage generation request
AudioSpeech or audio processing request
ErrorFailed request (no charge applied)

Common Use Cases

Auditing Spending

  1. Set the date range to the billing period you want to review
  2. Sort by Cost to find the most expensive requests
  3. Expand individual entries to see the full billing breakdown
  4. Use the Model Name filter to see spending per model

Debugging Issues

  1. Filter by Type: Error to see all failed requests
  2. Use Request ID to look up a specific failed request reported by your application
  3. Check the Channel and Duration columns to identify latency or availability issues

Optimizing Costs

  1. Filter by a specific model to see how much it's costing you
  2. Compare Prompt and Completion token counts to find opportunities to reduce input size
  3. Check if switching to a cheaper model for certain tasks would save money using the Model Hub

For a higher-level view of your spending, visit the Billing page.