deepseek-v4.1POST

DeepSeek V4.1 Flash hits 100% on an agentic bench in DeepSeek Harness

A user reports the oMLX-quantized DeepSeek V4.1 Flash scored 100% on their agentic bench in DeepSeek Harness at default settings, while the same model in PI had a wider trajectory and lower ceiling.

Kev

@KevrsubArchived 8 days agoOriginally posted 2026.10.24

Kev @KevrsubOriginally posted 2026.10.24

Open original on X

I wanted to test what both did at default. On an agentic bench I use, DSH with Deepseek v4.1-Flash (jundots oMLX quant) was the first and only local model to date to score 100% on this bench. I didn't change anything with DSH or PI. Effort -50, reasoning budget of 7k. Same model with PI had a wider trajectory and lower capability ceiling. I was more focused on the bench results (quality) and DSH had a way longer wall clock time and higher tokens but I couldn't argue with the quality. Happy testing 😉

x.com/kevrsub/status/2113759334811500977Archived 2026.10.24 12:07