Cursor now ships with a built-in LLM router that picks Claude, GPT, or local Llama per file.
Quick Answer
Cursor 0.52 introduces an integrated LLM router that selects between Claude, GPT, or local Llama models based on file-specific criteria like language and task type.
Quick Take
This feature optimizes cost and performance for users, allowing power users to customize model selection per workspace.
Key Points
- Integrated router selects models based on language, length, and task class.
- Supports Claude, GPT, and local Llama models for cost-effective performance.
- Power users can override model selection on a per-workspace basis.
- Optimizes file handling by choosing the cheapest competent model.
- Enhances user experience by streamlining model selection process.
Article Excerpt
From source RSS / original summaryCursor 0. 52 ships an integrated router. It picks the cheapest competent model per file based on language, length, and task class. Power users can override per-workspace.
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.