Skip to content
Thread
Out
Recently unrolled
Topics
How it works
Unroll
Testing
1 unrolled thread about testing, each one readable on a single page.
Popular
Recent
Paweł Huryn
@PawelHuryn
So I tested GPT-6.1 Sol on real work: 2 repos, 105 planted bugs. Find and fix what you can. Unlike GPT-6 Sol, which was just a nerfed GPT-5.6 Terra, this one is real. The results: - GPT-6 Astra (max...
Summary
A developer tested GPT-6.1 Sol on finding and fixing planted bugs across two repositories, comparing it against other models. GPT-6.1 Sol achieved competitive results (44/105 bugs) at a fraction of the cost ($6.56) compared to more expensive alternatives, making it a cost-effective option for code analysis tasks.
Sep 29, 2026
2 tweets
♥ 2.4K
