Unverified50% confidenceBenchmarkExact time
Cursor has launched CursorBench 4.0 for evaluating agent performance on real-world, ambiguous, multi-file tasks, and the addition of long-horizon problems has caused scores to drop across all models.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/11/2026
First Seen
Valid until: 12/10/2026
Sources
Related Entities
Related Claims
UnverifiedCursor co-founder Michael Truel stated that Opus 4.8 outperforms previous Opus models at every effort level on CursorBench, with more efficient tool calls and fewer steps.73% similarUnverifiedCursor was reported to automatically downgrade models during peak periods.72% similarUnverifiedCursor提供Debug Mode，可生成假设、添加日志埋点、引导复现问题并分析根本原因71% similarUnverifiedCursor具有调试模式（Debug Mode）功能，基于ReAct框架的智能体工作流，可自动配置监控探针、生成假设并逐一测试70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/908254API
curl https://kongchang.com/api/v1/knowledge/claims/908254MCP
get_claim(id=908254)