Connect Cline to Open-Weight Models on Impossible Cloud Inference

Interactive demo

The demo below walks through the full setup and a short coding task. The steps that follow describe the same configuration in text form.

Prerequisites

• Visual Studio Code installed
• An Impossible Cloud inference account and API key
• The model ID you want to use, from the Impossible Cloud model catalogue

Step 1: Get your endpoint URL, API key and model ID

 You can find your endpoint URL in the Impossible Cloud console, and create an API key there. The model IDs are listed in the Impossible Cloud model catalogue. 

You need three values for the configuration in Cline:

• Base URL: the OpenAI-compatible endpoint, ending in /v1,
  for example https://eu-west-3.inference.impossibleapi.net/v1
• API key: your Impossible Cloud inference API key
• Model ID: the exact ID from the model catalogue, for example qwen3.8-27b

Step 2: Install Cline in VS Code

1. In VS Code, open the Extensions view.
2. Search for Cline and install the extension.
3. Open Cline from its icon in the activity bar.

Step 3: Choose "Bring my own API key"

1. On the How will you use Cline? screen, select Bring my own API key.
2. Click Continue.

No Cline subscription is required for this setup. Cline sends requests directly to your Impossible Cloud endpoint.

Step 4: Configure the OpenAI Compatible provider

On the Configure your provider screen, enter the following:

1. API Provider: select OpenAI Compatible.
2. Base URL: enter your Impossible Cloud endpoint URL, including /v1. Cline adds the rest of the path.
3. OpenAI Compatible API Key: paste your Impossible Cloud API key. Enter the key only, without a Bearer prefix. The key is stored locally in VS Code.
4. Model ID: select the model from the dropdown. Cline loads the list directly from your endpoint.

For details on all fields, see the Cline documentation on OpenAI-compatible providers.

Step 5: Set the model configuration

1. Expand Model Configuration.
2. Supports Images: enable this only if the model accepts image input.
3. Context Window Size: enter the context window of the model from the Impossible Cloud model catalogue.
4. Input Price / 1M tokens and Output Price / 1M tokens: enter the token prices from the model catalogue. Cline uses these values to show the cost of each task.
5. Reasoning Effort: choose how much the model reasons before answering. Higher effort gives more thorough answers but uses more tokens.
6. Click Continue.

All settings can be changed later in the Cline settings.

Step 6: Start coding with Plan and Act mode

1. Check that the model shown below the chat input is your Impossible Cloud model.
2. Switch to Plan mode and describe your task. Cline reads the relevant files and proposes a plan without changing anything.
3. Review the plan, then switch to Act mode. Cline writes the code and asks for your approval before running commands or changing files, depending on your auto-approve settings.