Inference-time scaling improves LLM answer quality by using more compute during inference. This article categorizes approaches and reviews recent papers, showing that major LLM providers rely on such techniques.
OpenAI
The Register guides readers on setting up a local AI coding assistant using the Qwen3.6-27B model to avoid rising costs of cloud APIs. The article covers inference setup, hyperparameters, and three agent frameworks: Claude Code, Pi Coding Agent, and Cline. It concludes that while local models are not yet replacements for frontier models, they are surprisingly capable for many tasks.
Alibaba/Qwen
Anthropic
OpenAI
Microsoft
Meta