Skip to main content
LLMPodium
排行榜
🏆 Overall LeaderboardComposite index across all flagship LLMs💻 Coding & SWE-BenchReal-world programming & software agent benchmark🧠 Deep ReasoningGPQA Diamond, AIME & complex logic🤖 Autonomous AgentsOSWorld, Toolathlon & multi-turn execution🔓 Open WeightsDeepSeek, Qwen, Llama & Mistral
竞技场
⚔️ Human Arena BattlesBlind pairwise Elo rankings (700+ models)⚖️ Side-by-Side CompareHead-to-head metric comparison of up to 4 models💰 Cost-per-Task CalculatorEstimate API economics across real workflows🎯 Model RecommenderFind the optimal model for your budget and speed
模型列表
📦 Model DirectoryDetailed profiles, context windows & pricing🏢 AI Labs & ProvidersOpenAI, Anthropic, Google, DeepSeek, Meta📊 Benchmark MatrixEvaluation methodologies and leader tables
nav.resources
📐 Podium MethodologyMathematical aggregation formula explained📰 Changelog & NewsDaily model updates and new evaluations✍️ Research & ArticlesIn-depth AI benchmarking whitepapers❓ Frequently Asked QuestionsCommon questions on scores and ranking📖 LLM GlossaryDefinitions of TTFT, TPS, Elo, MoE & tokens
🇺🇸 EnglishEN🇨🇳 中文ZH🇮🇳 हिन्दीHI🇪🇸 EspañolES🇫🇷 FrançaisFR🇸🇦 العربيةAR🇧🇩 বাংলাBN🇧🇷 PortuguêsPT🇷🇺 РусскийRU🇵🇰 اردوUR🇮🇩 Bahasa IndonesiaID🇩🇪 DeutschDE🇯🇵 日本語JA🇰🇷 한국어KO🇹🇭 ไทยTH🇮🇹 ItalianoIT
排行榜
🏆 排行榜
Overall Composite LeaderboardCoding & Software AgentsDeep Reasoning & MathAutonomous Agent SwarmsOpen Weights & Self-Hosted
⚔️ 竞技场 & Compare
Human Battle Arena (700+ Models)Side-by-Side Model CompareCost-per-Task CalculatorInteractive Model Finder
📦 Directories & Evals
Model Specifications DirectoryAI Labs & Cloud ProvidersBenchmark Methodologies
📖 Research & Info
Scoring MethodologyResearch Blog & ArticlesChangelog & New ModelsFrequently Asked QuestionsTechnical LLM GlossaryAbout LLMPodium
🌐 Language / Язык / 语言
🇺🇸English🇨🇳中文🇮🇳हिन्दी🇪🇸Español🇫🇷Français🇸🇦العربية🇧🇩বাংলা🇧🇷Português🇷🇺Русский🇵🇰اردو🇮🇩Bahasa Indonesia🇩🇪Deutsch🇯🇵日本語🇰🇷한국어🇹🇭ไทย🇮🇹Italiano
Explore Full Leaderboard →
  1. 首页
  2. /AI 提供商
  3. /Moonshot AI

Moonshot AI

Moonshot AI 的 16 个模型,按 Podium 分数排名。

Kimi K3
最佳模型
83.3
Podium 综合得分
52.1
平均分
3
开源权重
Kimi K3
83.3
1.0M ctx37 t/s$15/M out
83.3
开源
Kimi K3 Max
74.0
128K ctx50 t/s$3/M out
74.0
Kimi K2.5 Thinking
73.0
128K ctx50 t/s$3/M out
73.0
Kimi K2.5 Instant
72.0
128K ctx50 t/s$3/M out
72.0
Kimi K2 Thinking Turbo
72.0
128K ctx50 t/s$3/M out
72.0
Kimi K2 0905 Preview
71.0
128K ctx50 t/s$3/M out
71.0
Kimi K2 0711 Preview
71.0
128K ctx50 t/s$3/M out
71.0
Kimi K2.6
49.0
262.1K ctx40 t/s$4/M out
49.0
开源
Kimi K3 Low
48.3
131K ctx60 t/s$4/M out
48.3
Kimi K2.7 Code
41.8
262K ctx39 t/s$4/M out
41.8
开源
Kimi K2 5
36.0
131K ctx60 t/s$4/M out
36.0
Kimi K2 6 Non Reasoning
35.4
131K ctx40 t/s$4/M out
35.4
Kimi K2 Thinking
33.5
131K ctx40 t/s$4/M out
33.5
Kimi K2 5 Non Reasoning
30.1
131K ctx40 t/s$4/M out
30.1
Kimi K2 0905
24.0
131K ctx60 t/s$4/M out
24.0
Kimi K2
19.7
131K ctx60 t/s$4/M out
19.7
LLMPodium

The definitive open LLM benchmark aggregator and leaderboard. Continuous evaluations across 719+ models and 25+ benchmarks.

Leaderboards synced daily
排行榜
  • 排行榜
  • 💻 编程
  • 🧠 推理
  • 📐 数学
  • 🤖 智能体
  • 🔓 开源权重
  • ⚡ 速度
  • 💰 性价比
分析工具
  • 竞技场 (719+ models)
  • 模型对比 Side-by-Side
  • 基准测试 Matrix
  • 选模型
  • 模型列表 Directory
  • AI 提供商
  • Cost per Task Calculator
  • #1 Ranked Model Profile
技术资源
  • 新闻 Feed
  • 博客文章 & Research
  • Best AI Models 2026
  • Enterprise AI Use Cases
  • 评分方法
  • 术语表
  • 常见问题
  • 关于我们
© 2026 LLMPodium. 保留所有权利。·所有评测数据均定期从公开权威数据源同步。
llms.txtRSS FeedOpen API