Stanford: small local models match cloud LLMs on 98.6% of chat tasks

Stanford researchers found small local models (Qwen 3, Gemma 3, GPT-OSS, Granite 4.0) won or tied cloud LLMs (ChatGPT 5, Claude Sonnet 4.5, Gemini 2.5 Pro) on 98.6% of chat queries and 62.5% of reasoning tasks. The author argues this could make hyperscaler data-center investments obsolete.