BotConnector
ID
warning
Legacy managed-runtime reference. Command examples below describe the older desktop/managed llama.cpp preview and are legacy, not the current BCCLI v0.4 command surface. Use BCCLI and Local AI for current instructions.

API problems

Connection refused

  1. Is the server running? botconnector server status
  2. Start it: botconnector server start (botconnector server)
  3. Confirm port: default is 11435 — a custom --port changes every URL.

404 on model

The model field must equal the loaded model id:

powershell
botconnector ps --json

Use exactly that id in requests. A model downloaded but not loaded is not served.

401 / authorization errors

  • An --api-key was set at server start; send Authorization: Bearer <key> — Authentication.
  • Rotated the key? Update every client.

Wrong port / stale processes

  • botconnector ps shows what is actually running.
  • A previous instance may hold the port — stop it or use another port and update clients.

From another device fails

Localhost binds only to this machine. Remote access requires explicit binding + auth — see Network binding and Remote.

Malformed responses / hangs

  • Verify with a raw request:
bash
curl http://127.0.0.1:11435/v1/models
  • If curl works but a client fails, the client's base URL is likely wrong (missing /v1 is the most common error).
  • Streaming clients: confirm SSE handling — Streaming.