AnalysisAI ModelsNovember 18, 2026

DeepSeek V4.1 Flash 1017 and Kimi K3.1 Pro tested on LongTail-SWE

Read original source →reddit.com

DeepSeek V4.1 Flash 1017 scored 76% on LongTail-SWE, up from 57% for V4 Flash, while Kimi K3.1 Pro hit 77% to top open-weight models. The benchmark covers 202 real GitHub tasks in 13 less common programming languages.

1 source

More stories today

Open the live feed