One-shot LLM eval cases with a bilingual (zh/EN) showcase site: Vite + React + Tailwind v4 + GSAP, built with bun, deployed to GitHub Pages. Ships two Claude Fable 5 Max cases (Mythos Craft voxel game, pelican-by-the-sea SVG) with placeholder slots for GPT-5.1, Gemini 3 Pro, DeepSeek V3.2 and Qwen3 Max. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
1.6 KiB
1.6 KiB
NAGI BENCH
NAGI STUDIO 的 LLM 测评案例集:同一段提示词,不同模型,一次生成、不许返工,并排对比它们交出的可运行作品。
One-shot LLM eval cases by NAGI STUDIO: same prompt, different models, one attempt and no retries — a side-by-side record of the runnable artifacts they ship.
Live site: https://nagi-studio.github.io/nagi-bench/
结构 / Structure
outputs/<model-id>/<artifact> 模型产出原文件(HTML / SVG),按模型分目录
site/ 展示站点(Vite + React + Tailwind v4 + GSAP,bun 驱动)
site/src/data/cases.ts 案例定义:双语提示词、模型列表、产出文件映射
.github/workflows/deploy.yml GitHub Pages 自动部署
本地开发 / Development
cd site
bun install
bun run dev # 同步 outputs/ 并启动 http://localhost:5173/nagi-bench/
bun run build # 类型检查 + 产物构建(输出到 site/dist)
添加案例或模型 / Adding a case or model
- 把产出文件放进
outputs/<model-id>/。 - 在
site/src/data/cases.ts里登记:- 新模型:加入
MODELS(status: 'ran' | 'pending')。 - 新案例:加入
CASES(中英双语title/tagline/prompt)。 - 产出映射:在
OUTPUTS里把案例 id 对应到模型 id 与文件路径。
- 新模型:加入
- 推送到
main,GitHub Actions 自动构建并部署。
Drop the artifact into outputs/<model-id>/, register it in site/src/data/cases.ts (model in MODELS, case in CASES with bilingual prompts, file path in OUTPUTS), then push to main — GitHub Actions deploys automatically.