In the SWE-bench test, Devin was able to correctly resolve 13.86% of GitHub issues without any assistance, performing far better than GPT-4.
View Article on VentureBeat
AI,Business,Programming & Development,AI bug fixes,AI code,AI code discovery,AI Coding,AI Coding Assistant,AI copilot,AI for coding,AI programmer,AI software engineer,AI, ML and Deep Learning,category-/Computers & Electronics/Programming,category-/Science/Computer Science,coding,cognition,Cognition AI,Conversational AI,Devin,GitHub Copilot,NLP,Replit,StarCoder
AI Coding