The field of large language models (LLMs) is shifting towards a more sustainable and cost-efficient approach. Researchers are exploring ways to reduce the environmental footprint and financial costs of LLMs without compromising their performance. This is achieved through the development of smaller, locally deployable models that can handle everyday occupational tasks with strong and reliable results. The focus is on task- and context-aware sufficiency assessments, which prioritize organizational priorities over performance-maximizing benchmarks. Moreover, innovative methods such as vocabulary adaptation and end-to-end optimization are being proposed to improve the efficiency of LLMs. Noteworthy papers in this area include: Sustainability via LLM Right-sizing, which evaluates the performance and sustainability of various LLMs across different tasks and proposes a shift towards task-aware sufficiency assessments. Cost-of-Pass: An Economic Framework for Evaluating Language Models, which introduces a framework for evaluating language models based on their cost-effectiveness and provides insights into the progress of cost-efficiency in LLMs. From Large to Super-Tiny: End-to-End Optimization for Cost-Efficient LLMs, which proposes a three-stage pipeline for deploying cost-efficient LLMs. Optimizing LLMs for Italian: Reducing Token Fertility and Enhancing Efficiency Through Vocabulary Adaptation, which presents a novel method for optimizing English LLMs for the Italian language. Energy Considerations of Large Language Model Inference and Efficiency Optimizations, which analyzes the energy implications of common inference efficiency optimizations and provides insights for sustainable LLM deployment.