LLM Arena .skill

Make 50+ LLM providers compete. May the best model win.
让 50+ 大模型供应商内卷竞争,优胜劣汰!
npx github:Han-cy830/llm-arena
Claude Code Skill MIT License Node.js 16+ Python 3.10+
📊

10-Dimension Scoring

10 维评分

Token efficiency, speed, quality, consistency, error rate, cost-efficiency, latency stability, improvement trend, win rate, user satisfaction

Token效率、思考速度、回答质量、一致性、错误率、性价比、延迟稳定、进步趋势、胜率、用户满意度

♟️

Elo Rating

Elo 评级

Chess-level Elo rating system, K=32. Beat a strong opponent to gain more, lose to a weak one to lose more.

国际象棋级 Elo 评分系统,K=32。击败强敌涨分多,输给弱敌扣分多,排名公平公正

⚔️

Blind Battle

盲测对决

Inspired by Chatbot Arena. Hide model identity, let users vote for the better answer blindly.

借鉴 Chatbot Arena,隐藏模型身份让用户盲投,选出真正更优的回答

😊

Emoji Moods

emoji 心情

19 dynamic moods: Champion, Dark Horse, Overconfident, Losing Streak, Grinding Hard...

19种心情动态变化:😎王者、🦄黑马、🤡飘了、😢连败、🔥势头正猛、⚙️疯狂内卷...

🏆

Achievements

成就系统

19 achievements: First Blood, Penta Kill, Unstoppable, Giant Slayer, Comeback King...

19种成就:🩸首战告捷、💀五杀、🌪️势不可挡、👑逆袭之王、🗡️屠龙勇士、💎完美主义...

👑

User Veto Power

用户否决权

Your ranking, your rules. Force ranks, ban models, rate directly. Algorithm can't override you.

你的排名你做主!可以强制指定排名、封禁模型、直接打分。算法再强也强不过你的裁决

Supported Providers

支持供应商

50+
Claude Official DeepSeek GPT-4o Gemini OpenRouter Volcengine DashScope Kimi Zhipu GLM SiliconFlow AWS Bedrock GitHub Copilot Shengsuanyun MiniMax Nvidia Youyun +40 more...

Emoji Mood System

emoji 心情系统

😎
Champion
王者
😤
Determined
志在必得
🦄
Dark Horse
黑马
🔥
On Fire
势头正猛
⚙️
Grinding
疯狂内卷
🤡
Overconfident
飘了
🥳
Win Streak
连胜中
😢
Losing Streak
连败中
🦅
Reborn
涅槃重生
🐣
Newcomer
新来的
😴
Slacking
摸鱼中
🦥
Given Up
摆烂了
💪
Catching Up
奋起直追
😰
Under Pressure
压力山大
🧓
Veteran
老将
💀
Banned
已封禁

Achievement Hall

成就殿堂

🩸 First Blood首战告捷
🎩 Hat Trick帽子戏法
💀 Penta Kill五杀
🌪️ Unstoppable势不可挡
Legendary超神
🦾 Iron Man铁人
💎 Perfectionist完美主义
⏱️ Speed Demon极速恶魔
💰 Budget Saver省钱达人
👑 Comeback King逆袭之王
📊 Rock Steady稳如老狗
🗡️ Giant Slayer屠龙勇士
🏃 Marathoner马拉松选手
Zero Mistakes零失误
📈 Rising Star进步之星
🏆 Champion冠军
🌟 Monthly MVP月度MVP
Weekly Star周度之星
❤️ User's Favorite用户最爱

Let the LLMs Battle

让大模型们开始内卷吧

One command. 50+ providers. 10-dimension scoring. Elo ratings. Emoji moods. Achievements.

一条命令,50+ 供应商,10维评分,Elo评级,emoji心情,成就系统

GitHub   Install Guide 安装指南