GXSIO API reference

The pages in this section render the GXSIO API's OpenAPI schema with Fumadocs OpenAPI. Open an endpoint group to see its parameters, response schemas and code samples, or send a request from the playground.

The health checks and the deprecated V1 security endpoint are in the sidebar.

Base URLCopy Anchor Link

https://dev.api.graphxserver.io/api

Every path in this reference is relative to that URL, including its /api prefix. V1 paths start with another /api, so a V1 request has it twice. That's correct, not a typo.

AuthenticateCopy Anchor Link

Send an organization API key in the Authorization header. Create one by following Create and manage API keys.

curl "https://dev.api.graphxserver.io/api/api/ProductMockupRequest/GetSuppliersForCompany?companyname=Acme" \
  --header "Authorization: Bearer $GXS_API_KEY"

The key decides which company's data you see. A request for a record that belongs to another company fails with "does not belong to your company". Load the key from your secret store and keep it out of shell history, code and logs.

Every endpoint except the three health checks needs this header. Without it, the API answers 401 with "Missing Authorization header". The schema only declares authentication on the V2 endpoints, but the V1 endpoints need it too, so this reference marks them as requiring it.

V1 and V2Copy Anchor Link

V1 copies the contract of the older GraphxServer.io API so existing integrations keep working. Its paths are PascalCase, such as GetProductListForCustomer, and successful responses wrap their data in a tObject field.

{
  "tObject": [
    { "id": 1, "title": "PCNA" },
    { "id": 7, "title": "GraphXSource" }
  ]
}

The schema doesn't describe V1 response bodies, so the reference shows them as "Empty". Use the playground to see a real response.

V2 uses resource-style paths such as /v2/location-aliases/{id}. Only location aliases exist in V2 so far. If you're starting a new integration and V2 covers what you need, use V2.

Some V1 endpoints exist in the schema but aren't built yet. They answer 501 with the code GXSIO.ENDPOINT_NOT_IMPLEMENTED. In the development environment this includes GetAllProductLines and GetGraphXServerAPIToken.

ErrorsCopy Anchor Link

Errors come back as JSON problem details with a matching HTTP status.

{
  "type": "https://graphxserver.io/problems/unprocessable",
  "title": "Unprocessable request",
  "status": 422,
  "detail": "A product identifier is required",
  "instance": "/api/ProductMockupRequest/GetProductPartDetails",
  "code": "GXS.UNPROCESSABLE",
  "traceId": "5c8dc4d33be2623..."
}

Include the traceId when you report a problem. It lets the API team find the request in their logs.

The schema's description says V1 returns business errors as HTTP 200. That isn't what the development API did when this page was written. Missing parameters, unknown records and unbuilt endpoints all returned 400, 404, 422 and 501. Check the status code, and don't treat every 200 as a success until the API team confirms which behavior is intended.

Use the playgroundCopy Anchor Link

The playground sends real requests to the development API from your browser. Paste an API key into the authorization field, and use a key from a test organization. POST, PATCH and DELETE requests change data.

Endpoint summaries and descriptions are in Spanish because they come straight from the API's source code. This reference shows them unchanged.

Some query parameters, such as CustomerName and CompanyName, are marked required. In testing, GetProductListForCustomer returned the key's own products without CustomerName, so the key seems to decide the company. Send these parameters anyway until the API team says they can be dropped.

Live reference and schemaCopy Anchor Link

• Live Swagger UI
• Live OpenAPI JSON
• Schema snapshot used by these pages

The snapshot was downloaded from the development API on October 16, 2026. The live API can change before these pages are refreshed.

The downloaded schema has no servers entry, doesn't define the bearer scheme its V2 operations use, and doesn't declare authentication on V1. This reference fills those gaps when it renders the pages. The snapshot file keeps the schema exactly as the server returned it.

Refresh the referenceCopy Anchor Link

From the docs-internal repository root:

# Download the development schema and regenerate the API pages.
pnpm --filter fumadocs gxsio:openapi:refresh

# Regenerate pages from the checked-in schema without downloading it.
pnpm --filter fumadocs gxsio:openapi

Commit the schema snapshot and the generated pages together, and update the download date above. Builds read the local snapshot, so they don't need the API to be reachable.

The script reads the version from each tag name. A tag like v2 · Products becomes v2/products.mdx, and a tag without a version lands at the top of the API section next to Health. New pages appear at the end of their folder, after the ones listed in that folder's meta.json. Add them there to change the order, and add a title to scripts/gxsio-openapi.mjs if the tag name isn't sentence case.