Organization API
Read your organization's data from scripts and monitoring tools. All routes use GET and the base URL https://api.boardwalk.cloud/v1/org. Your key selects the organization; no organization ID is needed.
Key permissions
In the console, create a key with only the permissions your integration needs. For a read-only key, leave inference:write unchecked.
| Permission | Access |
|---|---|
inference:write | Run inference, tokenize, and read receipts. |
requests:read | Read all organization requests and receipts. |
deployments:read | Read organization deployment history and details. |
models:read | Read organization models and their configuration. |
audit:read | Read the full organization audit trail, including every member's events. |
Read scopes cover the whole organization, including other members' activity. Reads work at zero credits. To change permissions, replace the key and revoke the old one.
Endpoints
| Path | Permission |
|---|---|
/v1/org/requests | requests:read |
/v1/org/requests/{request_id} | requests:read |
/v1/org/receipts/{request_id} | requests:read or inference:write |
/v1/org/models | models:read |
/v1/org/models/{id} | models:read |
/v1/org/deployments | deployments:read |
/v1/org/deployments/{id} | deployments:read |
/v1/org/audit-events | audit:read |
/v1/org/audit-events/{id} | audit:read |
Requests and receipts
See the request reference and receipt reference for request parameters and response fields.
curl --get https://api.boardwalk.cloud/v1/org/requests \
-H "Authorization: Bearer $BOARDWALK_API_KEY" \
--data-urlencode 'status=server_error' \
--data-urlencode 'limit=100'Filter request status by success, client_error, or server_error. List entries have a state:
completed: reached a worker; a receipt records the outcome, including failures.refused: rejected before a worker, with zero cost and no receipt.running: available on a detail lookup when live tracking is present; excluded from lists.
A detail lookup can briefly return 404 before a record exists. Use receipt_url to fetch billing details. Prompts and completions are not returned.
Models and deployments
See the model inventory and deployment reference.
Model inventory includes pending, active, and disabled models. Use a returned id for detail reads and its slug for inference. Draft prices are not active rates.
Deployment filters accept running, succeeded, failed, or stopped. Details include actual phase names, timestamps, checks, and failure reasons. Start deployments in the console.
Audit events
See the audit event reference for filters and response fields.
Events name the actor, action, resource, and recorded before/after state. Filters include actor_user_id and resource_type. The audit trail covers recorded changes, not every API read.
Pagination
cursorstringoptional- Opaque next_cursor from the previous page, at most 256 characters. Keep all other filters unchanged.
limitintegeroptional- Page size from 1 to 200. Default: 50.
{ "object": "list", "data": [], "next_cursor": null }limit defaults to 50 and allows up to 200. Pass next_cursor as cursor with the same filters until it is null. Lists are newest first.
from and to are epoch milliseconds: inclusive start, exclusive end. Keep the end fixed during an export and deduplicate overlapping polls by record ID. Response timestamps are ISO 8601 UTC; money is returned as decimal strings in micro-dollars.
Errors and limits
Organization reads have their own request budget. Honor Retry-After on 429. See API errors for other statuses. A record outside your organization returns 404.
OpenAPI schema
Download the OpenAPI 3.1 schema for all organization response fields, filter values, and client generation. The schema is public.