COMPARATIVE ANALYSIS OF LLM-BASED AI ASSISTANTS FOR EFL WRITING INSTRUCTION: CHATGPT, CLAUDE, GEMINI, AND DEEPSEEK

Abstract

This study investigates the comparative capabilities of four prominent LLM-based AI assistants (ChatGPT, Claude, Gemini, and DeepSeek) in supporting English as a Foreign Language (EFL) writing instruction, particularly in providing explanations of key linguistic concepts. Employing a three-phase research design, the study examines how these AI tools assist EFL teachers in explaining grammar, vocabulary usage, sentence structure, and cohesion/coherence. In Phase 1, standardized prompts were developed based on Khan’s prompt engineering framework to ensure consistency across AI interactions. Phase 2 involved submitting these prompts to each AI system under controlled conditions using their free versions to reflect common educational constraints. In Phase 3, the AI-generated responses were analyzed thematically using Braun and Clarke’s framework, focusing on clarity, conceptual depth, organization, and instructional value. The findings reveal distinct pedagogical profiles: ChatGPT offers structured and practice-oriented explanations, Claude emphasizes conceptual depth and academic reasoning, Gemini employs accessible and engaging explanatory strategies, and DeepSeek provides concise explanations suitable for quick reference. This study provides practical guidance for EFL teachers in selecting AI tools based on instructional context and learner needs.

Ähnliche Arbeiten