We compare ChatGPT, Gemini, and Claude on the tasks teachers actually care about: generating lesson plans, giving writing feedback, and creating exam practice materials.
Three AI platforms now dominate the conversation in ELT: ChatGPT (OpenAI), Claude (Anthropic), and Gemini (Google). Each has genuine strengths for teachers — and each has limitations that matter in a classroom context. Here's what we found after testing all three on real teaching tasks.
On lesson planning, all three performed well, but with different strengths. ChatGPT tends to produce more structured, template-style lesson plans with clear timing and objectives. Claude produces more nuanced plans that feel closer to how an experienced teacher actually thinks — it's more likely to anticipate pedagogical issues and suggest alternatives. Gemini's strength is integration: if you're already in the Google ecosystem, it connects naturally with Docs, Slides, and Classroom.
For writing feedback, Claude is the standout performer. Its feedback is detailed, specific, and — crucially — kind in tone. It's less likely to produce the kind of blunt, discouraging comments that can demotivate language learners. ChatGPT is also strong here, especially with grammar. Gemini lagged behind on nuance, though it has improved significantly in recent months.
On exam material creation — gap fills, multiple choice, IELTS-style questions — ChatGPT is arguably the most reliable. It handles structured output formats consistently and produces well-calibrated difficulty levels when prompted correctly.
Our overall recommendation for most teachers: start with Claude or ChatGPT, depending on whether you want more nuanced prose output (Claude) or more reliable structured formats (ChatGPT). Use Gemini if you're deeply embedded in Google Workspace. All three offer free tiers that are sufficient for most classroom use.