Read request counts, token totals, and costs for a time window, grouped by model and API key name. Requires usage:read; get a key in the console first if needed.
GET /open/v1/requests/summarySummary parameters
All parameters are optional URL query parameters.
| Parameter | Type | Description |
|---|---|---|
startTime |
integer | Inclusive start time in Unix seconds; defaults to 24 hours before endTime |
endTime |
integer | Inclusive end time in Unix seconds; defaults to the current time |
model |
string | Exact model name |
keyName |
string | Exact API key name |
outcome |
string | succeeded, interrupted, or failed |
The maximum window is 31 days. Both times must be positive, and the start cannot be later than the end. Split longer periods into smaller windows and add their totals. Since both boundaries are inclusive, start each subsequent window one second after the previous end to avoid counting boundary records twice.
curl --fail-with-body --silent --show-error \
--connect-timeout 5 --max-time 15 --get \
https://api.xiu.ai/open/v1/requests/summary \
-H "Authorization: Bearer $XIU_OPEN_API_KEY" \
--data-urlencode "startTime=1794355200" \
--data-urlencode "endTime=1794441599"Summary fields
{
"startTime": 1794355200,
"endTime": 1794441599,
"requests": 1520,
"succeeded": 1490,
"interrupted": 6,
"failed": 24,
"promptTokens": 2840000,
"completionTokens": 612000,
"cost": "12.481250",
"currency": "USD",
"groups": [
{
"model": "gpt-5",
"keyName": "production",
"requests": 1520,
"succeeded": 1490,
"interrupted": 6,
"failed": 24,
"promptTokens": 2840000,
"completionTokens": 612000,
"cost": "12.481250"
}
]
}| Field | Type | Description |
|---|---|---|
startTime, endTime |
integer | Actual time bounds, including resolved defaults, in Unix seconds |
requests |
integer | Total requests, equal to succeeded + interrupted + failed |
succeeded, interrupted, failed |
integer | Counts by outcome, with the same meanings as request records |
promptTokens, completionTokens |
integer | Total input and output tokens, excluding failed requests |
cost |
string | Exact total cost with six decimal places |
currency |
string | Currency, currently USD |
groups |
array | Groups sorted by model, then keyName |
Each group contains the string fields model and keyName, followed by its own requests, succeeded, interrupted, failed, promptTokens, completionTokens, and cost, with the same types as above. To report by model or key alone, add the corresponding groups.
A period without requests returns zero totals and an empty groups array. For reconciliation, use identical time bounds and filters for records and summaries. After reading every page, the decimal sum of individual cost values equals the summary cost. The requestId filter is available only for individual records, not summaries.
For the individual records behind a total, use List request records. See errors and rate limits for retry guidance.