AI usage
AI usage shows every token your organization spent, broken down by model, feature, and individual record, with a per-record drill-down.
AI usage is the most detailed view of what your AI is doing. Where Cost insights tells you what you spent, AI usage tells you which model spent it, on what feature, for which record. Go to Insights → AI usage to open the page. Use it to chase down the source of a cost spike and to attribute spend to the right widget, project, or conversation.
What gets tracked
Every AI call in Informly produces a usage record. The page rolls those records up along three dimensions you can pivot on.
| Dimension | Examples |
|---|---|
| Model | GPT-4, Claude, embeddings model, text-to-speech voice. |
| Feature | Chat reply, document embedding, voice synthesis, skill execution. |
| Record | A specific document, project, conversation, or user. |
Pivoting on model tells you which model is the most expensive in your stack. Pivoting on feature tells you whether chat or embeddings is driving the bill. Pivoting on record tells you the exact item to investigate.
Per-record drill-down
Every document, project, and conversation has its own AI usage page.
Open the record
Go to the document, project, or conversation in question. A Usage tab sits in the header or sidebar.
Read the timeline
The tab shows usage over time for that single record — tokens spent on it, the models that processed it, the features it triggered.
Spot the outliers
A conversation that consumed 10x the tokens of its neighbors is worth opening. A document whose embedding cost keeps reappearing was probably re-uploaded several times.
The per-record view is the fastest way to attribute a spike. If Cost insights shows a jump on Tuesday, open the day on AI usage, sort conversations by tokens descending, and the top three rows are almost always the cause.
Export the report
Click the menu in the top right and choose Export CSV to download the report. The export contains one row per usage record with the model, feature, record type and ID, token count, and cost. It respects whatever filters and date range you have applied.
The CSV is the right format if you want to feed Informly usage into a finance tool or build your own internal dashboard.
How usage relates to billing
Plan quotas cover a baseline of AI usage. Anything above that quota is on-demand and draws from your wallet.
Check your quota
Go to Billing → Plan to see this month's included tokens and how many you've used. See Usage limits.
Cross-reference with AI usage
The total tokens on the AI usage page for the current month should match the usage shown on the plan page.
Fund the wallet if needed
If you're trending over quota and want to keep going, top up the wallet. If you're consistently over, consider upgrading the plan instead.
What's next
Cost insights
Cost insights break down what you're spending on AI by feature, widget, and project, with day-over-day deltas so you can catch spikes early.
Plans
Informly plans bundle documents, widgets, conversations, and AI usage at a fixed monthly or annual price, with custom pricing for larger teams.