GonkaGate Cline OpenAI Compatible Setup

Connect the Cline VS Code extension to GonkaGate through the OpenAI Compatible provider. Set the provider values, click Verify, then run one short prompt before you trust a longer coding session.

Set the provider values

Open Cline settings from the gear icon, then set:

API Provider: OpenAI Compatible
Base URL: https://api.gonkagate.com/v1
API Key: gp-your-api-key
Model: <model-id-from-get-v1-models>

If the model does not appear in the picker, enter the exact GonkaGate model ID manually from GET /v1/models.

Before you start

The Cline extension is already installed in VS Code.
You already have a GonkaGate API key in gp-... format.
You need the narrow Cline provider setup, not a broader SDK, migration, or raw API walkthrough.

Verify the provider connection

Click Verify in the Cline settings panel first.

After that succeeds, run one short prompt in chat:

Reply with exactly: Cline connected to GonkaGate

That confirms the provider, base URL, API key, and model path before you start a real coding session.

Cline-specific rules

Use the exact model ID returned by GonkaGate. Do not assume OpenAI default model names.
Cline lets you choose the model from a picker or enter it manually. Manual entry is the safer path if the model is missing from the picker.
Leave Model Configuration alone unless you intentionally need custom context-window, max-output-token, image-support, or computer-use hints for a known model.
Verify checks the connection path, but one real chat prompt is still the safer canary before a long session.

Common first failures

If you see	What it usually means	What to do
`Verify` fails or you get a connection error	The Base URL is wrong or the machine cannot reach the endpoint	Recheck `https://api.gonkagate.com/v1`, then confirm network access to the endpoint
`401 invalid_api_key`	The API key is missing, invalid, or copied from the wrong environment	Recheck the key and Authentication and API Keys
`404 model_not_found`	The model ID is stale or invalid	Switch to an exact model ID from GET /v1/models
The model connects but behaves oddly	Custom `Model Configuration` hints changed the expected behavior	Reset the custom hints and retry with a plain text request
`429 insufficient_quota`	The prepaid USD balance is too low for this request	Check Pricing before retrying
`429 rate_limit_exceeded`	The request was throttled	Retry with bounded backoff and use Rate Limit Handling for the full policy

Official references

Last verified: 2026-03-13 against the official Cline OpenAI-compatible provider docs and Cline CLI docs.

Cline Setup