Skip to main content
GET
/
v1
/
analytics
/
computer
/
usage
Computer Usage Analytics
curl --request GET \
  --url https://api.perplexity.ai/v1/analytics/computer/usage \
  --header 'Authorization: Bearer <token>'
{
  "data": [
    {
      "start_time": 1746057600,
      "end_time": 1746144000,
      "count": 350,
      "by_category": [
        {
          "category": "paid",
          "count": 250
        },
        {
          "category": "promo",
          "count": 80
        }
      ]
    }
  ],
  "has_more": false,
  "next_page": null
}
This endpoint requires an organization analytics API key, generated by an org admin from the Perplexity web app — not a regular Sonar API key. See Computer Analytics API for setup, time-window semantics, and pagination.

Authorizations

​
Authorization
string
header
required

Bearer authentication header of the form Bearer <token>, where <token> is your auth token.

Query Parameters

​
dataset
enum<string>
required

The dataset to query.

Available options:
credit_usage,
connectors,
artifacts,
skills,
spaces,
workflows,
task_durations
​
start_time
integer
required

Window start in unix seconds (UTC), inclusive. Snapped down to the bucket grid, so the first bucket can include usage from before this time.

Required range: x >= 0
​
end_time
integer

Window end in unix seconds (UTC), exclusive. Defaults to now; future values are capped at now. The window may span at most 90 days.

Required range: x >= 0
​
bucket_width
enum<string>
default:1d

Bucket size. Buckets align to the UTC grid.

Available options:
1d,
1h
​
limit
integer

Buckets per page. 1d: default 7, max 31. 1h: default 24, max 168.

​
page
string

Opaque pagination cursor from a previous response's next_page. Valid only with the same query parameters it was issued for.

​
user_email
string<email>

Restrict results to a single member of your organization. Emails that don't match a current member return a generic 400.

Response

Bucketed usage for the requested window. Buckets without data carry count 0; the analytics store syncs periodically, so a zero in a recent bucket can mean the data hasn't synced yet.

​
data
object[]
required

Every bucket in the page's window, in chronological order. Buckets without data carry count 0 and an empty breakdown.

​
has_more
boolean
required

Whether more buckets exist beyond this page. False also when the window's remaining buckets are past the current data frontier.

​
next_page
string | null

Cursor for the next page; pass as the page parameter with otherwise identical query parameters.