> ## Documentation Index
> Fetch the complete documentation index at: https://docs.openference.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Cline

> Configure Cline in VS Code.

# Cline

In Cline settings:

* API Provider → **OpenAI Compatible**
* Base URL: `https://api.openference.com/v1`
* API Key: your Openference API key
* Model ID: `GLM-5.2` (or another ID from the [Models page](https://openference.com/models))

## Set the context window size

Cline (and forks like Roo Code and Kilo Code) default OpenAI-compatible
providers to a **128k context window** and start auto-compacting the
conversation well before that (\~100k tokens). To use the full context of
larger models, open the model configuration for the OpenAI Compatible
provider and set **Context Window Size** (and **Max Output Tokens**, where
available). The Context Window Size is the safe **input** budget — the model's
total context wall minus Max Output Tokens, with a small \~2% safety buffer.
Models with a smaller output cap get a larger input budget.

| Model             | Context Window Size | Max Output Tokens |
| ----------------- | ------------------- | ----------------- |
| GLM-5.2           | 850000              | 128000            |
| DeepSeek-V4-Pro   | 850000              | 131072            |
| DeepSeek-V4-Flash | 850000              | 131072            |
| Qwen3 Code        | 915000              | 65536             |
| Qwen3.7 Plus      | 915000              | 65536             |

Current values for every model are also advertised as `context_length` and
`max_output_tokens` in [`GET /v1/models`](/api-reference/models).