Update docs/cline-cli/cli-reference.mdx

docs/cline-cli/cli-reference.mdx

@@ -265,6 +265,20 @@ TASK SETTINGS
 
        mode   Starting mode (act/plan)
 
+CONTEXT WINDOW CONFIGURATION
+       The context window setting varies by provider:
+
+       Ollama
+              cline config s ollama-api-options-ctx-num=32768
+
+       LM Studio
+              cline config s lm-studio-max-tokens=32768
+
+       Other providers (Anthropic, OpenRouter, etc.)
+              The context window is defined per model in the model metadata.
+              It is not a user-settable field via CLI. Models have their own
+              built-in context limits that Cline uses automatically.
+
 NOTES & EXAMPLES
        The cline task send and cline task new commands support reading from
        stdin, enabling powerful pipeline compositions:

docs/running-models-locally/lm-studio.mdx

@@ -87,6 +87,16 @@ After loading your model in the Developer tab, configure these settings:
 2. **KV Cache Quantization**: Leave unchecked (critical for consistent performance)
 3. **Flash Attention**: Enable if available (improves performance)
 
+#### Configure context window in Cline
+
+By default, Cline requests a specific context window size from LM Studio. To use a custom context window:
+
+1. Open Cline settings
+2. Scroll to the model configuration section
+3. Set your desired context window size in the "Context Length" field
+
+Ensure the value matches or is less than the context length configured in LM Studio's Developer tab.
+
 #### Quantization Guide
 
 Choose quantization based on your RAM:

docs/running-models-locally/ollama.mdx

@@ -84,6 +84,26 @@ Other capable models include:
 - `mistral-small` - Good balance of performance and speed
 - `devstral-small` - Optimized for coding tasks
 
+### Configure context window size
+
+By default, Cline requests a 32,768 token context window from Ollama. If you've configured a custom context size in your Ollama model, Cline will override it unless you explicitly set the context window in Cline's settings.
+
+To use a custom context window size:
+
+1. Open Cline settings
+2. Scroll to the model configuration section
+3. Set your desired context window size in the "Context Length" field
+
+If you want Ollama to use the context size defined in your model's Modelfile (via `num_ctx`), set the context length in Cline to match that value.
+
+Example Modelfile with custom context:
+```
+FROM qwen3:30b
+PARAMETER num_ctx 100000
+```
+
+Then set Cline's context length to `100000` to match.
+
 ### Important Notes
 
 -   Start Ollama before using with Cline