CrowdinQA

OpenAI Usage Insights

Real consumption pulled directly from the OpenAI Admin API using the key configured in Settings → LLM Engines → OpenAI.

Loading…
OpenAI Admin API key required.
OpenAI's usage endpoints require an Admin key (sk-admin-...), not a project key. Generate one in platform.openai.com → Organization → Admin keys and save it in Settings → LLM Engines → OpenAI Admin (Usage API only). The regular OpenAI provider entry is for chat/completions and won't work for this dashboard.
Fetching OpenAI usage…
Total Cost
—
 
Requests
—
 
Input Tokens
—
 
Output Tokens
—
 
Cached Tokens
—
cache rate —
Cost / Request
—
cost / 1k tok —
Images Generated
—
Speech Chars (TTS)
—
Transcription mins
—
Vector Stores
—
Code-Interp. Sessions
—
Moderations
—
Daily Cost — stacked by endpoint
Cost by Endpoint
Requests Heatmap — hour × day-of-week Switch to Hourly bucket to populate
Prompt Cache Hit Rate completions only
Top Models by Requests
Input vs Cached vs Output Tokens — per Model
Cost by Project
Project Requests Tokens Cost
Activity by API Key
API Key Requests Input tok. Cached tok. Output tok.
Images by Size
Audio Split
TTS chars
0
STT seconds
0
Storage & Sessions
Vector Store Bytes
—
—
Code Interp. Sessions
—
sandbox runs
Cost Breakdown by Line Item
Line item Cost (USD) Share
Daily Activity Detail
Date Requests Input tok. Cached tok. Output tok. Cost Cost bar