Get Runs Chart Data
Retrieve aggregated run data optimized for charting over long time ranges.
Authorizations
Service Account Token authentication. To authenticate API requests:
-
Create a Service Account Token:
- Go to the Cerebrium Dashboard and open the API Keys page
- Click Create Service Account, name it (e.g., "GitHub Actions CI/CD"), choose an expiry date, and click Create
- Copy the token generated for the desired service account
-
Use the Token: Include the service account token in the Authorization header of API requests:
Authorization: Bearer <your-service-account-token> -
Best Practices:
- Create separate service accounts for different environments (dev, staging, prod)
- Store tokens securely as secrets in consuming applications or workflows
- Set appropriate expiry dates and rotate tokens regularly
- Never commit tokens to source control
For CI/CD integration examples, see the CI/CD documentation.
Query Parameters
Start of the time range, in RFC3339 format.
End of the time range, in RFC3339 format.
Maximum number of individual runs to return when not aggregating. Must be > 0. Default: 100.
Number of runs to skip when not aggregating. Must be >= 0. Default: 0.
Filter runs by status. Valid values: pending, proxyQueued, containerQueued, processing, success, failure, cancelled.
Filter runs by HTTP status code.
Time bucket for aggregated results. Valid values: minute, hour, day.
Set to true to return aggregated counts instead of individual runs. Time ranges longer than 24 hours are aggregated automatically.
Only include runs with a startup time of at least this many milliseconds.
Only include runs with a startup time of at most this many milliseconds.
Only include runs with a response time of at least this many milliseconds.
Only include runs with a response time of at most this many milliseconds.
IANA timezone used for time bucketing (e.g. America/New_York). Default: UTC.
Response
Run data for charting, either individual runs or aggregated time buckets.
Aggregated counts per time bucket, when aggregating. Each item includes timeGroup, startTime, endTime, request counts (totalRequests, successfulRequests, failedRequests, cancelledRequests, proxyQueued, containerQueued, processing), successRate (number), and timing statistics in milliseconds (avg, p50, p90, p99, min, and max for runtime and response time, plus avgColdstartTimeMs and avgQueueTimeMs). Omitted when not aggregating.
The time bucket size used for aggregation. Omitted when not aggregating.
Individual runs, when not aggregating. Each item has id, projectId, modelId, podName, coldstartTimeMs (integer), totalQueueTimeMs (integer), runtimeMs (number), totalResponseTimeMs (number), statusCode (integer), createdAt, and updatedAt fields. Omitted when aggregating.
Pagination token for the next page. Omitted when there are no more results.
The time range covered by the aggregated results. Omitted when not aggregating.
Total number of runs matching the query.