Usage analytics

The Usage page in your organization dashboard shows how your team uses Hull Gateway: spend, requests, tokens and who is active.

What's included

Usage covers requests that go through Hull Gateway. Requests a member sends to another provider with their own key don't appear here.

Views

Tabs at the top of the page switch between views.

View	What it shows	Plan
AI usage	Summary metrics, the AI adoption score and breakdown charts	Teams and Enterprise
Overview	Feature adoption and AI adoption summary side by side	Enterprise
Feature adoption	Which Hull features the organization has set up, with recommendations	Enterprise

The AI usage view opens with:

Use the side panel to change what the page shows.

Control	Options
Scope	My Usage, the organization, and any sub-organizations
Period	Today, Yesterday, Past Week, Past Month, Past Year
Granularity	Hour, Day, Week, Month
Cost	Cost (what you paid) or Estimated Market Cost
Metric	Cost, Requests, Tokens, Input Tokens, Output Tokens, Error Rate, Avg Latency, Avg Generation Time, Cost / Request, Tokens / Request, Cache Hit Ratio, Output / Input Ratio
Dimension	Feature, Model, Mode, User, Provider, Project
Filters	Include or exclude specific values of a dimension

You can export the usage table to CSV.

Requests are tagged with a project name so you can group usage by Project. Hull resolves the name in this order:

project.id in .hull/config.json in the project (or the older .hullcode/config.json)
The repository name from the origin Git remote

For example, a remote of git@github.com:example-co/example-repo.git gives the project name example-repo.

To set the name yourself, create .hull/config.json:

{
  "project": {
    "id": "my-project"
  }
}

The AI usage view also shows the organization's AI adoption score. See AI adoption score.