Press Space to continue
DeepSeek V4 Pro achieves 90.5% on ARC-AGI reasoning benchmark
Press Space to continue
DeepSeek V4 Pro from @deepseek_ai on ARC-AGI (Verified): - ARC-AGI-1: 90.5%, $0.18/task - ARC-AGI-2: 61.3%, $0.60/task Scores were similar across low, high, and max reasoning, and its top scores were comparable to DeepSeek V4 Flash.
open on x · 1 viewViews and clicks count once per visitor per day. Clicks are opens of this post on X.