Skip to main content

Usage Log

The Usage Log records what your organization spent, one row per day and one column per source — the part of Abacus.AI the usage came from.

Find it under Setting and Permissions → Usage. Organization administrators only. Amounts are shown in dollars.

Columns are ordered with the largest spend first, so the order changes as your usage does, and a column appears only once it has been used. Group by User splits the same totals by person, and Download CSV exports whatever you are currently looking at.

Dates are in UTC, and usage can take up to an hour to appear.


What each column covers​

ColumnWhat it covers
UIOrdinary chat. Also the default for anything without its own column below.
Abacus AI AgentThe agent: research, writing code, building and deploying apps.
Personal AgentsAssistants your users have saved and named, and the jobs they run.
Scheduled TasksWork that runs on a schedule.
Custom Router UIChats answered by a routing rule you configured.
Embedded ChatbotA chatbot you embedded in your own site, used by your visitors.
Media StudioEverything inside a Media Studio session, including what it produces.
Media GenerationImages, video and speech produced outside Media Studio.
Document ProcessingReading and extracting text from documents you upload.
Powerpoint GenerationBuilding PowerPoint decks.
Abacus AI DesktopThe desktop app — chat, agent mode, edits and autocomplete.
Abacus AI CoWorkCoWork sessions in the desktop app.
AbacusAI BotThe Abacus AI Bot desktop assistant.
OfficeThe Microsoft Office add-in, from Word, Excel, PowerPoint or Outlook.
Browser ExtensionThe Abacus.AI browser extension.
Voice AgentsLive spoken conversations.
Audio ListenerThe background listener that takes notes from audio.
Abacus ClawUsage from Abacus Claw.
API UsageCalls made to Abacus.AI with an API key.
RouteLLM APICalls to the RouteLLM API endpoint.
Custom Router APIAPI calls answered by a routing rule you configured.
Abacus AI Agent LLM APIModel calls made through the agent's LLM API.
Deployment TokenPredictions served from your deployments using a deployment token.
AI EngineerThe AI Engineer assistant, including code help in notebooks.
LLM MonitoringAutomated quality checks on models you have deployed.
Chatbot MetricsAutomated scoring of your chatbots' answers.
Supercomputer LLMsModel usage inside a cloud machine.
Supercomputer HostingThe running cost of that machine and its disk.
Claw LLMs, Claw HostingModel usage, and running cost, for Claw.
Database HostingStorage for databases we host for you.
Supporting ServicesBehind-the-scenes work — recalling context, summarising, checking answers.
Email NotificationsThe summary and notification emails the product writes for you.
RefundsAmounts credited back to you, shown as a negative figure.

RouteLLM usage is billed to the model that answered, so it appears under that model rather than as a line of its own.