DeepSeek V4.1 Flash 在 ARC-AGI 验证集上超越 V4 Flash
DeepSeek 的 V4.1 Flash 跑分出来了,ARC-AGI-2 拿下 72.9%,比上一代高了 11.5 分,不过每任务成本贵了两倍多,看你怎么权衡。
DeepSeek 发布的 V4.1 Flash 在 ARC-AGI 验证集上公布成绩:ARC-AGI-2 得分 72.
9%,单任务成本 $0.13;
ARC-AGI-1 得分 94.5%,单任务成本 $0.
07。相比 V4 Flash 的最好成绩,V4.
1 Flash 在 ARC-AGI-1 上高出 5.5 分,在 ARC-AGI-2 上高出 11.
5 分。不过代价是单任务成本比 V4 Flash 高约 250%。
成绩来自 ARC Prize 官方验证。
DeepSeek V4.1 Flash from @deepseek_ai on ARC-AGI (Verified):
- ARC-AGI-2: 72.9%, $0.13/task - ARC-AGI-1: 94.5%, $0.07/task
DeepSeek V4.1 Flash beats V4 Flash's best scores by 5.5 points on ARC-AGI-1 and 11.5 on ARC-AGI-2, but costs about 250% more per task. https://t.co/DDOiSntuR0
This report is categorized under AITOP's "Models" desk, focusing on China's AI industry and domestic foundation model ecosystem. Within our cross-regional analytical framework, it directly maps to US-China foundation model benchmarks and commercialization comparison and AI developer toolchains, coding agents, and engineering workflows. Key entities and technical topics under continuous tracking include DeepSeek, ARC-AGI, reasoning models, open-source models.
Source attribution: curated from ARC Prize (@arcprize).