About the model settings used for automatic compaction

[Decade-qiu]

Question

I have enabled auto-compaction, but I couldn't find any model setting options for auto-compaction in the documentation.

It seems to be using the currently selected model for compaction right now, but this is not very user-friendly for Copilot users, as it consumes one request quota. I hope there is a compaction model option, so i can use the cheap one to compact my context.

[github-actions]

This issue might be a duplicate of existing issues. Please check:

• [FEATURE]: opencode could automatically use different models based on task type #8456: Feature request for using different models based on task type (addresses the core request for cheaper models for specific operations)
• Feature Request: Configurable context limit and auto-compaction threshold #8140: Configurable context limit and auto-compaction threshold settings
• Missleading documentation or incorrect behaviour about small_model #8609: Documentation and configuration of small_model for lightweight tasks
• [FEATURE]: Use a cheap model to name sessions #4133: Feature request to use a cheaper model for lightweight operations like session naming

Feel free to ignore if none of these address your specific case.