←Back to NewsAI News/LlmsrepoLlmsOpen SourceOpen-Source Jeff Turns Tiny Models Into 22ms Local Decision APIsJeff packs zero-shot classification into 0.8B parameter models that decide in 22ms, trained on one workstation GPU in about two hours.SourceAlphaSignalPublishedSep 29, 2026, 2:25 PMAuthorAlphaSignal NewsroomRead1 min readJeff packs zero-shot classification into 0.8B parameter models that decide in 22ms, trained on one workstation GPU in about two hours.Reporting is indexed from AlphaSignal. Rights remain with the original publisher and cited sources.Read original report ↗Next readsAlphaSignal · modelSupersonic Labs Ships Julia-1, a Tiny 144M Router That Classifies Without RetrainingAlphaSignal · modelLEGAL-BERT Hits 833K Monthly Downloads Helping Developers Ditch Costly AI ModelsAlphaSignal · modelDuoNeural's Cyber-Ornith Puts a 9B Hacking Agent on Consumer GPUs
