Real-SWE's top-scoring AI coding agent reportedly failed over 60% of the time
The New Stack reports that Real-SWE tests AI models on private, proprietary codebases.
Posts from X
TN 
The New Stack@thenewstackAI coding agents stumble on private codebases: Real-SWE benchmark tests models on proprietary code, and the top scorer still failed over 60% of the time.
- likes: 1
- replies: 0
- bookmarks: 1
- reposts: 0
TN The New Stack@thenewstackAI's best coding agent fails 60% of the time — and the data backs it up
- likes: 0
- replies: 2
- bookmarks: 0
- reposts: 0