←Back to NewsAI News/DevelopmentnewsDevelopmentPost TrainingCognition's SWE-2 Beats GPT-5.6 Sol at 64% Lower CostCognition's new coding model matches frontier systems while costing up to 70% less, driven by an RL recipe that trains every effort level at once.SourceCognitionPublishedSep 10, 2026, 3:21 PMAuthorAlphaSignal NewsroomRead1 min readCognition's new coding model matches frontier systems while costing up to 70% less, driven by an RL recipe that trains every effort level at once.Reporting is indexed from AlphaSignal. Rights remain with the original publisher and cited sources.Read original report ↗Next readsAlphaSignal · repoSoup Trains Llama-3.1-8B on a 4 GB Laptop GPU for FreeAlphaSignal · repoJianying Headless Lets Coding Agents Build Editable Video Drafts From JSONPydantic · newsPydantic AI Adds Jev to Cut Classification Latency 6x Without Generating Tokens
