ainewsblitz.com

Breaking

Kwaipilot's 32B Open Coding Model KAT-Dev Reports 62.4% on SWE-Bench Verified, Rivaling Far Larger Rivals

Kwaipilot's 32B Open Coding Model KAT-Dev Reports 62.4% on SWE-Bench Verified, Rivaling Far Larger Rivals
  • Software Dev & Coding
  • Open Source
  • Foundation Models

Kwaipilot says its compact, open-weight coding model KAT-Dev-32B resolves 62.4% of tasks on SWE-Bench Verified, a result the developer positions as fifth among open-source models regardless of size while running at a fraction of the parameter count of larger competitors.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.

$20
Read this article
$29/month
Unlimited — all 27,934 articles, the full archive, and comprehension quizzes
Save 72%
$98/year
≈ $8.17/month
Unlimited, billed once a year