Gemini Tops ChatGPT, Claude in Blind Essay Test
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- StudyArena analyzed 6,851 blind student votes across AI models for essay tasks, finding Gemini won with a 39.6% choice rate ahead of Claude (31.8%) and ChatGPT/OpenAI (29.2%) as of August 2026.
- Gemini led writing feedback at 41.7%, Claude led assignment planning at 43.2%, and ChatGPT led research work at 39.3% — meaning each model has a distinct specialty beyond the headline ranking.
- Students preferred longer answers: the selected response was 37% longer on average than alternatives, and the longest response won 47.7% of decisive writing comparisons.
- Higher reasoning/effort settings did NOT produce higher choice rates in writing tasks; StudyArena recommends normal or low effort for prose, reasoning that extra reasoning adds repetition that hurts essays.
- The August 2026 data is drawn from current model families — GPT-5.6 Sol, Claude Opus 5, and Gemini 3.1 Pro — not older generations still cited elsewhere.
- StudyArena recommends using AI as an editor rather than a ghostwriter, with essay-type-specific prompts that ask models to diagnose rather than rewrite.
Why it matters: Blind methodology removes brand loyalty as a confound, making Gemini's 39.6% edge a verdict on output quality rather than reputation. The specialty split — Gemini for editing (41.7%), Claude for planning (43.2%), ChatGPT for research (39.3%) — gives students a concrete multi-model workflow, though a single Gemini subscription covers the most common task of improving a finished draft.
Ask SkimNews
