WHAT NEW IN X10.29 THU

X 上的 AI 合集档案 · 22 个合集 · 58,925 条

deepseek-v4.1POST

V4.1 Flash 多项跑分超 GPT-5.6 Sol:HLE 42.6、DeepSWE 77.0

发帖称 DS-V4.1-Flash 在 HLE 得 42.6、Terminal-Bench 42.3、DeepSWE 77.0,多项跑分超过 GPT-5.6 Sol,且这还不是完整版。

Jaysen♨️
@_jayyy18_13k10 天前收录原推发布于 2026.10.17
Jaysen♨️ @_jayyy18_13k原推发布于 2026.10.17
在 X 上打开原推

what a surprise from deepseek 👀 DS-V4.1-Flash is beating GPT-5.6 Sol on several benchmarks and it's not even the full version 42.6 on HLE, 42.3 on Terminal-Bench and 77.0 on DeepSWE that’s a serious showing for a flash model deepseek might have something special here

x.com/_jayyy18_13k/status/2111401389431914553采集于 2026.10.18 16:42