排行榜

能力评测站内热度

外部权威评测，保留原始分数与来源

综合编程 Agent 与工具推理与知识写作与对话语言行业视觉图像视频

综合编程前端 · React 前端 · HTML 设计稿转代码仓库级修复算法题

算法题LiveCodeBench：固定时间窗口内的新竞赛题，Pass@1 客观正确率。

第 1 名

O4-Mini (High)OpenAI

LiveCodeBench Pass@1

77.10%

领先第 2 名

+4.58

来源

LiveCodeBench2025-08-02

算法题排名26

model= /

名次模型LiveCodeBench Pass@1

01

O4-Mini (High)

O4-Mini (High)

77.10%

02

Gemini 2.5 Pro 06 05

Gemini-2.5-Pro-06-05

72.52%

03

O3 (High)

O3 (High)

71.76%

03

O4-Mini (Medium)

O4-Mini (Medium)

71.76%

05

DeepSeek R1 0528

DeepSeek-R1-0528

68.70%

06

Gemini 2.5 Pro 05 06

Gemini-2.5-Pro-05-06

67.18%

07

OpenReasoning-Nemotron-32B

OpenReasoning-Nemotron-32B

65.65%

08

O4-Mini (Low)

O4-Mini (Low)

62.60%

08

O3-Mini-2025-01-31 (High)

O3-Mini-2025-01-31 (High)

62.60%

08

Grok-3-Mini (High)

Grok-3-Mini (High)

62.60%

11

OpenCodeReasoning-Nemotron-1.1-32B

OpenCodeReasoning-Nemotron-1.1-32B

61.37%

12

Gemini 2.5 Flash 05 20

Gemini-2.5-Flash-05-20

59.54%

13

O3-Mini-2025-01-31 (Med)

O3-Mini-2025-01-31 (Med)

56.49%

13

Gemini 2.5 Flash 04 17

Gemini-2.5-Flash-04-17

56.49%

15

Qwen 3 235B A22B

Qwen3-235B-A22B

55.73%

16

O3-Mini-2025-01-31 (Low)

O3-Mini-2025-01-31 (Low)

51.15%

16

Claude-Sonnet-4 (Thinking)

Claude-Sonnet-4 (Thinking)

51.15%

18

Claude-Opus-4 (Thinking)

Claude-Opus-4 (Thinking)

48.85%

19

Claude Sonnet 4

Claude-Sonnet-4

40.46%

20

Claude Opus 4

Claude-Opus-4

39.69%

有 7 天前的可比快照后显示周变化。