Cursor ships native GLM 5.2

Natively hosting a massive 744B open-weights model like GLM 5.2 is a major win for Cursor users looking for cheaper, long-context alternatives to Claude 3.5 Sonnet. By eliminating the friction of API keys and custom proxy configurations, Cursor makes high-end 'vibe coding' accessible to the masses.

• –Zero-config 1M context: Developers can now leverage GLM 5.2's massive 1-million-token context window directly in Chat or Composer without managing custom endpoints or keys.
• –Cost-effective performance: GLM 5.2 offers coding and reasoning capabilities that rival frontier models like Claude 3.5 Sonnet and GPT-5, but at a fraction of the inference cost.
• –Simplified multi-model workflows: Integrating GLM 5.2 directly in the settings makes it easy to chain planning tasks in Sonnet with execution/refactoring runs in GLM 5.2.
• –Avoids custom naming conflicts: Built-in integration solves previous unauthorized API errors caused by Cursor's internal routing intercepting manual OpenRouter/Fireworks custom endpoints.