Abstract: This paper focuses on the construction of an educational vertical large language model (LLM) for the mechanics “101 Plan”: The “AIM” (AI for mechanics education). In response to the learning requirement of students “constructing dual brains (human brain and AI)” in the digintel (digital and intelligent) era, AIM integrates technologies such as knowledge graphs, virtual simulation, generative AI, electronic teaching-plan databases, etc. It realizes full-link AI empowerment in teaching, learning, assessment, evaluation, management, and research, by building an integrated system of four graphs covering knowledge, competence, value, and quality; developing 15 major teaching and learning function modules such as engineering modeling; breaking through technical bottlenecks for “text-to-quantitative graph”; creating virtual simulation scenarios such as hypersonic wind tunnels; forging a hybrid teaching model of classroom, MOOCs, knowledge graphs and LLM; constructing a digital teaching material community; and so on. AIM contributes to the improvement of talent cultivation achievements such as mechanics competitions, and provides a demonstration for promoting the transformation of mechanics higher education from “knowledge impartment” to “competence construction” and the digital construction of more disciplines.