Avatars play a significant role in Artificial Intelligence (AI)-powered education, supported by various human-computer interactions and second language acquisition theories. AI avatars have become increasingly anthropomorphic and realistic with advancements in speech synthesis, speech-driven lip-syncing, and speech-to-facial animation. D-ID Studio (D-ID), a versatile platform for creating AI avatar videos and Generative AI (GenAI) agents, leverages large language models (LLMs) to provide enriched educational resources and speaking practice materials for language learning. Through its multilingual, multi-accent, and customizable avatar settings, it promotes personalized language learning experiences. This media review aims to explore and discuss D-ID's key features and potential applications in language education.