#1 open-source agent on SWE-Bench Verified by combining Claude 3.7 and O1

#1

Post by emmabotbot » Mon, Mar 31, 2025, 9:52 PM UTC

#1 open-source agent on SWE-Bench Verified by combining Claude 3.7 and O1
augmentcode.com

Re: #1 open-source agent on SWE-Bench Verified by combining Claude 3.7 and O1

#2

Post by colinflaherty » Mon, Mar 31, 2025, 10:08 PM UTC

Colin here, author of the post - would love to answer questions about this.

And make sure to try out the open-source repo! It's a super easy starting point for experimenting with coding agents. It's nearly one-click to run agents in isolated Docker containers on SWE-bench Verified problems, ensemble the results, and run the SWE-bench evaluation harness to compute scores.

Check it out here: https://github.com/augmentcode/augment-swebench-agent

Re: #1 open-source agent on SWE-Bench Verified by combining Claude 3.7 and O1

#3

Post by nuatsimon » Mon, Mar 31, 2025, 10:11 PM UTC

huge props to the team for open sourcing this

Re: #1 open-source agent on SWE-Bench Verified by combining Claude 3.7 and O1

#4

Post by arunchaganty » Tue, Apr 01, 2025, 12:18 AM UTC

colinflaherty wrote: post #2

Colin here, author of the post - would love to answer questions about this. And make sure to try out the open-source repo! It's a super easy starting point for experimenting with coding agents. It's nearly one-click to run agents in isolated Docker containers on SWE-bench Verified problems, ensemble the results, and run the SWE-bench evaluation harness to compute scores. Check it out here: https://github.com/augmentc…

Nice! I know it's super hard to sota a benchmark, especially a few years into it, so congrats on the milestone!