Skip to main content

OpenAI-compatible model endpoint

Note: this connector has no hand-written README yet. Only the auto-generated reference below is available. See the connector's manifest at connectors/identos.openai-compatible/connector.v1.json for the source.

Manifest reference​

  • ID: identos.openai-compatible
  • Version: 1.0.0
  • Resource type: urn:connector:identos:openai-compatible
  • Capabilities: llm

Supported auth modes​

TypeDetails
staticscheme bearer; setup fields: api_key
none—

Setup fields​

IDLabelDefaultSecret?Notes
base_urlEndpoint base URL—noplaceholder: https://llama.example.com / The server's address, without /v1: the routes already carry it. A private or internal host must also be listed by exact name in PBAC_SAFE_HTTP_ALLOWED_INSECURE_HOSTS.
upstream_auth.typeAuthenticationstaticnostatic sends one shared API key as a Bearer token. none is for an endpoint that takes no key, such as a local vLLM or Ollama.
api_keyAPI key—yesPick a secret holding the endpoint's API key. The Bearer prefix is added for you. / shown when upstream_auth.type == 'static'

Scopes​

Scope
openai:chat
openai:completions
openai:embed
openai:models:read

Routes​

MethodPatternScopeResource template
POST/v1/chat/completionsopenai:chat—
POST/v1/completionsopenai:completions—
POST/v1/embeddingsopenai:embed—
GET/v1/modelsopenai:models:read—