Cline through Model Gate

1. Open the Cline settings.
2. Select an OpenAI-compatible provider.
3. Set Base URL:

https://api.model-gate.com/v1

1. Set API key:

mg_live_...

1. Enter a canonical model ID or enabled alias from /v1/models.
2. Save and send a test prompt.

Test request

curl https://api.model-gate.com/v1/chat/completions \
  -H "Authorization: Bearer mg_live_..." \
  -H "Content-Type: application/json" \
  -d '{"model":"ch-46s","messages":[{"role":"user","content":"Reply with OK"}]}'

If Cline exposes separate model context/input/output limits, use context_window, max_input_tokens, and max_output_tokens from the corresponding authenticated /v1/models object (or the Context, Max input, and Max output values shown on the authenticated /models page). A value of 0 means the limit is not configured in Model Gate; leave that manual client setting unset instead of guessing.