←Back to NewsAI News/Open SourcenewsOpen SourcePost TrainingAi2's AstaBrief 8B Writes Cited Research Reports 3.5x Faster in One PassAi2 open-sources an 8B Qwen3-based model that generates cited scientific reports 3.5x faster than Claude-powered pipelines, with full training data.SourceAi2PublishedOct 2, 2026, 3:35 PMAuthorAlphaSignal NewsroomRead1 min readAi2 open-sources an 8B Qwen3-based model that generates cited scientific reports 3.5x faster than Claude-powered pipelines, with full training data.Reporting is indexed from AlphaSignal. Rights remain with the original publisher and cited sources.Read original report ↗Next readsBespoke Labs · repoBespoke Labs' Nimble Beats a 27B Model at Classification Using Just 9BAlphaSignal · paperRadixArk's Miles v0.1 Cuts Trillion-Parameter RL Sync Time by 86%Hugging Face · newsHugging Face and Liquid AI's LFM2.5 Jumps 12 Points Across Four Agent Harnesses
