Skip to main content

leaderboard.cn

Explore

Frontier Models
Historical Trends
Local Models
Compare Models
Ranking Studio

Library

Models
Pricing
Organizations
Benchmarks
Data sources

leaderboard.cn

Benchmark details loading

Loading page content

About Ranking methodology Changelog

v0.1.26.a85cb1563d03浙ICP备09076226号-7

Back to benchmarks

Arena Agent Net ImprovementAgents

Official website

Results for 31 models collected from 2 public sources.

#ModelScore

1

Claude Fable 5Anthropic · High mode

13.9%

2

GPT-5.6 SolOpenAI · xHigh mode

10.9%

3

Claude Opus 4.8Anthropic · Thinking mode

9.4%

4

GPT-5.5OpenAI · xHigh mode

8.3%

5

Claude Opus 4.7Anthropic

8.2%

6

Claude Sonnet 5Anthropic · High mode

8.0%

7

GLM-5.2Zhipu AI · Max mode

6.9%

8

GPT-5.4OpenAI · High mode

6.7%

9

Claude Opus 4.6Anthropic

6.5%

10

Grok 4.5xAI

5.0%

31 models total