💻 Coding · llm-stats.com
LiveCodeBench
Continuously refreshed competitive programming problems immune to data contamination.
31
جانچے گئے ماڈلز
DeepSeek V4 Pro
سرکردہ ماڈل
93.5%
بہترین اسکور
DeepSeek
فراہم کنندہ
| # | ماڈل | LiveCodeBench |
|---|---|---|
| 1 | DeepSeek V4 Pro DeepSeek | 93.5% |
| 2 | Gemini 3 Pro Google | 91.7% |
| 3 | DeepSeek V4 Flash DeepSeek | 91.6% |
| 4 | DeepSeek V3.2 DeepSeek | 83.3% |
| 5 | Grok 4 Heavy xAI | 79.4% |
| 6 | Kimi K3 Moonshot AI | 74.7% |
| 7 | GPT-5.5 OpenAI | 74.3% |
| 8 | GPT-5.6 Terra OpenAI | 74% |
| 9 | GPT-5.3 Codex OpenAI | 74% |
| 10 | GPT-5.6 Luna OpenAI | 74% |
| 11 | MiniMax M3 MiniMax | 74% |
| 12 | GPT-5.6 Sol OpenAI | 73.7% |
| 13 | MiMo V2.5 Pro Xiaomi | 73.3% |
| 14 | Gemini 3.1 Pro Google | 72.7% |
| 15 | GLM-5.2 Z.ai | 71.3% |
| 16 | Claude Fable 5 Anthropic | 70% |
| 17 | Claude Opus 5 Anthropic | 70% |
| 18 | Gemini 3.6 Flash Google | 69.7% |
| 19 | Muse Spark Meta | 69.7% |
| 20 | Qwen3.6 Plus Alibaba | 69.7% |
| 21 | Gemini 3.5 Flash Google | 69.3% |
| 22 | Qwen3.7 Max Alibaba | 69% |
| 23 | Claude Opus 4.8 Anthropic | 67.7% |
| 24 | Grok 4.5 xAI | 67.7% |
| 25 | Nemotron 3 Ultra 550B NVIDIA | 67% |
| 26 | Hunyuan Hy3 Tencent | 66.7% |
| 27 | Kimi K2.7 Code Moonshot AI | 66.3% |
| 28 | Muse Spark 1.1 Meta | 63.3% |
| 29 | Inkling Thinking Machines | 63.3% |
| 30 | Claude Sonnet 5 Anthropic | 58.7% |
| 31 | Claude Opus 4.6 Anthropic | 58.3% |
بینچ مارک ڈیٹا llm-stats.com سے مرتب۔ مزید جانیں طریقہ کار میں۔