context.vn

aa Agentic Index

113 models evaluated

#ModelProviderTypeScore
1GPT-5.5OpenAIClosed74.1%
2Claude Opus 4.7 (Adaptive)AnthropicClosed71.3%
3Gemini 3.5 FlashGoogleClosed70.3%
4GPT-5.4OpenAIClosed68.0%
5Claude Opus 4.6 (Adaptive)AnthropicClosed67.6%
6MiMo-V2.5-ProXiaomiClosed67.4%
7DeepSeek V4 Pro (Max)DeepSeekOpen67.2%
8GLM-5.1Z.AIOpen67.0%
9DeepSeek V4 Pro (High)DeepSeekOpen66.7%
10Qwen3.7 MaxAlibabaClosed66.6%
11GLM-5-TurboZ.AIClosed66.1%
12Kimi K2.6Moonshot AIOpen66.0%
13Grok 4.3xAIClosed65.9%
14Qwen 3.6 Max (preview)AlibabaClosed64.8%
15Claude Opus 4.7AnthropicClosed64.6%
16Claude Opus 4.6AnthropicClosed64.2%
17GLM-5Z.AIOpen63.1%
18Qwen3.6-27BAlibabaOpen62.9%
19MiMo-V2-ProXiaomiClosed62.8%
20DeepSeek V4 Flash (High)DeepSeekOpen62.3%
21Muse SparkMetaClosed62.0%
22Qwen3.6 PlusAlibabaClosed61.7%
23Claude Sonnet 4.6AnthropicClosed61.6%
24MiniMax M2.7MiniMaxOpen61.5%
25DeepSeek V4 Flash (Max)DeepSeekOpen61.3%
26GLM-5V-TurboZ.AIClosed61.1%
27GPT-5.3 CodexOpenAIClosed60.5%
28GPT-5.2OpenAIClosed60.2%
29Claude Opus 4.5 ThinkingAnthropicClosed59.6%
30Claude Opus 4.5AnthropicClosed59.2%
31Gemini 3.1 ProGoogleClosed59.1%
32Kimi K2.5 (Reasoning)Moonshot AIClosed58.9%
33Kimi K2.5Moonshot AIOpen58.9%
34GPT-5.4 miniOpenAIClosed58.9%
35MiMo-V2-OmniXiaomiClosed58.6%
36Qwen3.6-35B-A3BAlibabaOpen58.3%
37GPT-5.2-CodexOpenAIClosed56.5%
38Qwen3.5 397B (Reasoning)AlibabaOpen55.8%
39Hy3 PreviewTencentOpen55.7%
40GLM-4.7Z.AIOpen55.0%
41GPT-5 (high)OpenAIClosed54.6%
42Qwen3.5-27BAlibabaOpen54.6%
43Qwen3.5 397BAlibabaOpen53.3%
44Mistral Medium 3.5 128BMistralOpen53.2%
45Qwen3.5-122B-A10BAlibabaOpen53.0%
46Gemini 3 ProGoogleClosed52.0%
47GPT-5.1OpenAIClosed51.3%
48GPT-5.1-Codex-MaxOpenAIClosed50.7%
49GPT-5.1-CodexOpenAIClosed50.7%
50Grok 4.1 Fast (Reasoning)xAIClosed49.3%
51GPT-5.4 nanoOpenAIClosed47.6%
52MiMo-V2-FlashXiaomiOpen47.3%
53GPT-5 (medium)OpenAIClosed45.8%
54Qwen3.5-35B-A3BAlibabaOpen44.1%
55Qwen3 MaxAlibabaClosed43.0%
56GLM-4.6Z.AIOpen42.9%
57Trinity-Large-PreviewArcee AIOpen42.6%
58Trinity-Large-ThinkingArcee AIOpen42.6%
59Grok 4xAIClosed41.5%
60Gemma 4 31BGoogleOpen40.9%
61Command A+CohereOpen40.9%
62DeepSeek V3.2DeepSeekOpen39.8%
63Grok 4 Fast (Reasoning)xAIClosed39.5%
64Claude 4 SonnetAnthropicClosed39.2%
65K-ExaoneLG AI ResearchClosed38.1%
66Ling 2.6 FlashInclusionAIOpen38.1%
67GPT-OSS 120BOpenAIOpen37.9%
68o3OpenAIClosed36.1%
69Grok Code Fast 1xAIClosed35.6%
70Gemini 3 FlashGoogleClosed35.0%
71Grok 4.1 FastxAIClosed33.0%
72Gemini 2.5 ProGoogleClosed32.7%
73Gemma 4 26B A4BGoogleOpen32.1%
74DeepSeek V3.1DeepSeekOpen31.9%
75o1OpenAIClosed31.1%
76GPT-OSS 20BOpenAIOpen27.6%
77GPT-4.1OpenAIClosed27.3%
78Mistral Small 4 (Reasoning)MistralOpen25.9%
79Mistral Small 4MistralOpen25.9%
80Gemini 3.1 Flash-LiteGoogleClosed25.7%
81GPT-4.1 miniOpenAIClosed25.1%
82Sarvam 105BSarvamOpen24.7%
83Kimi K2Moonshot AIClosed24.3%
84Nemotron 3 Nano Omni 30B A3BNVIDIAOpen23.9%
85Mistral Large 3MistralClosed21.7%
86GLM-4.5-AirZ.AIClosed21.0%
87DeepSeek-R1DeepSeekOpen20.8%
88DeepSeek V3.1 (Reasoning)DeepSeekOpen18.9%
89Gemini 2.5 FlashGoogleClosed15.0%
90Mistral Medium 3MistralClosed13.7%
91Solar Pro 2UpstageClosed12.7%
92Sarvam 30BSarvamOpen11.5%
93Mistral Large 2MistralClosed10.2%
94DeepSeek V3DeepSeekOpen8.8%
95Nemotron 3 Nano 30BNVIDIAOpen8.5%
96GPT-4oOpenAIClosed8.4%
97Granite-4.0-1BIBMOpen7.6%
98Llama 4 MaverickMetaOpen7.2%
99Claude 3 HaikuAnthropicClosed7.0%
100Gemma 4 E4BGoogleOpen6.9%
101Gemma 4 E2BGoogleOpen6.9%
102Exaone 4.0 1.2BLG AI ResearchOpen6.8%
103Granite-4.0-H-1BIBMOpen6.5%
104Llama 3.1 405BMetaOpen6.3%
105GPT-4.1 nanoOpenAIClosed5.8%
106Llama 4 ScoutMetaOpen5.2%
107Granite-4.0-H-350MIBMOpen4.9%
108Nova ProAmazonClosed4.7%
109Granite-4.0-350MIBMOpen4.4%
110Nemotron Ultra 253BNVIDIAOpen3.8%
111Gemma 3 27BGoogleOpen3.5%
112Exaone 4.0 32BLG AI ResearchOpen1.4%
113Phi-4MicrosoftOpen0.0%