BotConnector

Developer API Reference

Base URL: https://api.botconnector.id/v1

Authentication

Authorization: Bearer bc_live_...

API keys are created and revoked from Workspace.

GET /v1/models

Returns model IDs available to the authenticated account. A request without a valid key returns an authentication error.

POST /v1/chat/completions

OpenAI-compatible Chat Completions surface. Use a canonical model ID from GET /v1/models.

OpenAI Responses — GPT-6 Luna

GPT-6 Luna uses the native OpenAI Responses API on the Web App/BCCLI path for eligible paid plans. The Luna route is not exposed as a generic route for ordinary API-key clients.

Live surface:

  • POST /v1/responses
  • WebSocket /v1/responses
  • POST /v1/responses/input_tokens
  • POST /v1/responses/compact
  • GET /v1/responses/{response_id}
  • DELETE /v1/responses/{response_id}
  • POST /v1/responses/{response_id}/cancel
  • GET /v1/responses/{response_id}/input_items
  • POST /v1/conversations
  • GET|POST|DELETE /v1/conversations/{conversation_id}
  • GET|POST /v1/conversations/{conversation_id}/items
  • GET|DELETE /v1/conversations/{conversation_id}/items/{item_id}

Responses supports native streaming, prompt caching, previous_response_id, WebSocket stream_id, proxy mid-turn steering, BCCLI automatic compaction, and background retrieve/resume. Stateful resources are checked against the BotConnector owner before they are accessed through the shared OpenAI project.

Hosted endpoint boundary

Not yet part of the public hosted contract:

  • OpenAI Files and Vector Stores
  • OpenAI Embeddings
  • Rerank
  • OpenAI Audio, Realtime, and GPT-Live
  • OpenAI fine-tuning, Batch, and administration resources

Unsupported routes return an explicit error instead of silently falling back to another provider.

Account surfaces

Workspace provides API-key management, usage views, model catalog, integration guides, and account billing state. Internal account/admin routes are not part of this public reference.