pachote/ai-writing-tools-comparison-2026

A concise repo that benchmarks and compares AI writing tools (ChatGPT, Anthropic Claude, and local open-source models) side‑by‑side. It’s intended for researchers, engineers, product managers and technical writers who need reproducible, comparable evaluations of generation quality, cost, latency and safety. Key features: - standardized prompt set and test corpora for multiple writing tasks (summarization, rewriting, long-form, instruction following) - scripts to run models (API wrappers and local model launch/config) - automated metrics (ROUGE, BLEU, BERTScore), cost/latency logging and optional human-eval workflow - result aggregation, plotting and export; CI-ready benchmarks and reproducibility notes - configuration for API keys, model parameters, and licensing/usage guidance No marketing claims; focuses on transparent, repeatable comparisons.

このRepoのdescription / GitHub topics / READMEのAI要約に出現したAI関連キーワードをカテゴリ別に表示。各バッジは該当カテゴリの詳細ランキングへリンクします。