Open-weight

Studio

Includes only open-weight models, combining reasoning, coding, agent, multimodal, and long-context performance.Updated 07/22, 03:19

#ModelScore

1

GLM-5.2Zhipu AI

100.0

2

Hunyuan 3Tencent

95.3

3

Kimi-K2.7-CodeMoonshot AI

95.1

4

DeepSeek-V4-ProDeepSeek

93.5

5

Kimi-K2.6Moonshot AI

93.3

6

MiniMax M3MiniMax

90.0

7

GLM-5.1Zhipu AI

89.7

8

DeepSeek-V4-FlashDeepSeek

86.6

9

Qwen3.5-397B-A17BAlibaba

86.6

10

LongCat-2.0Meituan

85.8

109 models total

Home

Open-weight

Studio

Includes only open-weight models, combining reasoning, coding, agent, multimodal, and long-context performance.Updated 07/22, 03:19

#ModelScore

1

GLM-5.2Zhipu AI

100.0

2

Hunyuan 3Tencent

95.3

3

Kimi-K2.7-CodeMoonshot AI

95.1

4

DeepSeek-V4-ProDeepSeek

93.5

5

Kimi-K2.6Moonshot AI

93.3

6

MiniMax M3MiniMax

90.0

7

GLM-5.1Zhipu AI

89.7

8

DeepSeek-V4-FlashDeepSeek

86.6

9

Qwen3.5-397B-A17BAlibaba

86.6

10

LongCat-2.0Meituan

85.8

109 models total