Skip to main content

​
Purpose

See every AI request on your website: how many there were, what they cost and whether they worked. Use it to control your budget and to find errors.

​
AI Usage

AI Usage shows how much AI was used.
Path: T3AF > AI Usage
  • Request count – how many AI requests in the chosen period.
  • Tokens – how much text was sent and received (AI services bill by tokens).
  • By extension – which AI extension used the AI (AI Assistant, AI Chatbot and others).
  • By feature – for example seo.meta_description.
  • Time range – day, week or month.
AI Usage with totals, filters and the request log Use it to plan your budget and spot unusual spikes. Trends are also on the Dashboard.

​
AI Logs

AI Logs shows each single request.
Path: T3AF > AI Logs
  • Time, user, extension and feature
  • Provider, model and tokens
  • Success or failure
AI Logs with entry counts, filters and log entries Use it to find out why a request failed.

​
Scheduler & CLI

Background tasks and commands of AI Foundation. See Scheduler & CLI.
Path: T3AF > Scheduler & CLI AI Foundation Scheduler & CLI with Quick Reference and Scheduled Tasks
On your live website, the TYPO3 Scheduler must run every minute (a cron job on the server). Ask your hosting provider or developer.
If a change does not show up, clear the TYPO3 caches on the command line (or with Flush all caches in the backend).

​
OpenAI org statistics (optional)

Set openai_admin_api_key in Extension Configuration for organization-level usage charts. This is not the chat API key. See Configuration.

​
Privacy

How much is logged depends on the provider’s Logging privacy setting and on the limits in AI Permissions. Check these before you store full prompts and answers.

​
Weekly admin habit

  1. Go to AI Universe → AI Foundation → AI Usage and compare with last week.
  2. Go to AI Logs and look for repeated errors (same user, same feature).
  3. Errors don’t go away? Send the log details to Support.

​
When logs show high usage

  • Check AI Features — bulk tasks may need a cheaper model
  • Review group limits in AI Permissions
  • Ask editors if a script or loop triggered many requests

​
When logs show failures