deepseek-v4.1IMAGE

DeepSeek V4.1-Flash-1017 open weights and benchmark comparison vs Opus 5

Terminal-Bench 4.0 rose from 31.2 to 42.3 and DeepSWE hit 77.0 vs Opus 5's 74.0, though HLE and terminal tasks still trail; outputs are shorter and faster, with a notable jump in 3D generation.

WquGuru

@wquguruArchived 10 days agoOriginally posted 2026.10.17

DeepSeek V4.1-Flash-1017 open weights and benchmark comparison vs Opus 5 (image 1 of 3)

DeepSeek V4.1-Flash-1017 open weights and benchmark comparison vs Opus 5 (image 2 of 3)

DeepSeek V4.1-Flash-1017 open weights and benchmark comparison vs Opus 5 (image 3 of 3)

WquGuru @wquguruOriginally posted 2026.10.17

Open original on X

🎉 DeepSeek真的变了，以前从来不加班，刚刚悄然放出 V4.1-Flash-1017，HuggingFace上已经可以下载。这次实战能力提升明显： Terminal-Bench 4.0：31.2 → 42.3 ProgramBench：20.3 → 29.1 其中已经跑赢 Opus 5的部分： DeepSWE：77.0 vs 74.0 Automation-Bench：60.5 vs 50.3 Apex Bench：72.1 vs 65.8 当然也别吹过头，纯推理和终端任务，Opus 5 还是明显领先： HLE：42.6 vs 56.3 Terminal-Bench：42.3 vs 51.8 更有意思的是，分数涨了，输出反而更短、跑得更快，性价比比上一代 4.1 更高。比较明显提升的是 3D 能力，几句话、很少的素材就能从零做出完整项目。

This post quotes another post: x.com/deepseek_ai/status/2111375640717033777

x.com/wquguru/status/2111401069733687548Archived 2026.10.18 16:42