←Back to NewsAI News/BenchmarksnewsBenchmarksReasoningDeepSeek V4 Pro Hits 90% on ARC-AGI-1 but Stalls on Harder PuzzlesDeepSeek V4 Pro hits 90.5% on ARC-AGI-1 and 61.3% on ARC-AGI-2, but extra reasoning barely moves the needle on abstract puzzles.SourceARC PrizePublishedAug 21, 2026, 8:41 PMAuthorAlphaSignal NewsroomRead1 min readDeepSeek V4 Pro hits 90.5% on ARC-AGI-1 and 61.3% on ARC-AGI-2, but extra reasoning barely moves the needle on abstract puzzles.Reporting is indexed from AlphaSignal. Rights remain with the original publisher and cited sources.Read original report ↗Next readsEpoch AI · newsGPT-6 Astra Cracks a Decade-Old Voting Theory Problem Nobody Could SolveVals AI · newsVals AI's MysteryMechanism Benchmark Tests If AI Can Rediscover 222 Scientific LawsEpoch AI · newsOpenAI's GPT-6 Astra Cracks Epoch's Hardest Math Benchmark in 14 Months
