Operational

Analytics

Requests, tokens, latency and errors across the selected period, by model and by key.

Requests93,52012.4%last 30 days
Input tokens34.3M62% of total
Output tokens21M38% of total
Mean latency623 mstime to last token
Success rate98.75%0.4%1,167 failed

Usage

Requests over the last 7 days

Share of requests

Proportion of traffic served by each model.

  • Claude Sonnet 54%
  • DeepSeek V4.1 Flash2%
  • GLM 5.22%
  • GLM 5.3 Flash1%
  • GPT 5.6 Luna1%
  • DeepSeek V4 Pro1%

Response status

7,276 responses across the period.

  • 200Success67.6%
  • 429Rate limited8.8%
  • 500Provider error11.8%
  • 504Timeout11.8%

Usage by model

Requests, latency and spend per model.

Usage by API key

Request volume attributed to each active key.