Open Inventory / Docs
EndpunkteAiAiImage models

List image-generation models enabled for this deployment

Die API läuft auf deiner Anwendungsinstanz. Schemas und Feldnamen stammen unverändert aus der OpenAPI-Spezifikation. Browsergebundene Aufrufe benötigen eine Sitzung auf der Instanz.

GET
/ai/image-models

Requires the ai scope. Only allowlisted models with configured provider credentials are returned.

Authentifizierung

bearerAuth
headerAuthentifizierungBearer <inv_ token>

Header-Parameter

X-Organization-ID?string

Organization membership selected for this request. User-bound credentials may select any active membership; standalone tokens may only repeat their pinned organization ID.

Formatuuid

Antwortinhalt

Available image-generation models and the deployment default

application/json
  1. response
models*array<>
defaultModelId*string|null
Beispielanfrage
curl -X GET 'https://inventory.example.com/api/v1/ai/image-models'
Beispielantwort
{  "models": [    {      "id": "string",      "provider": "openai",      "model": "string",      "label": "string"    }  ],  "defaultModelId": "string"}

List Replicate counting models enabled for this deployment GET

Requires the ai scope. The same catalog is consumed by the web app and iOS app.

Match one photographed object to existing inventory items POST

Requires direct inventory.read permission and the ai scope/ai.use permission. The uploaded image is transient and is not attached to an item. The server describes the dominant object, shortlists a bounded inventory catalog, and visually reranks up to five advisory matches. Clients must let the user review the result instead of treating it as an exact identifier.