context.vn

apex Agents Aa

27 models evaluated

#ModelProviderTypeScore
1Gemini 3.5 FlashGoogleClosedgemini-3-5-flash
2Kimi K3Moonshot AIClosedkimi-3
3GPT-5.6 TerraOpenAIClosedgpt-5-6-terra
4GPT-5.5OpenAI · Closed37.7%
5GPT-5.6 LunaOpenAI · Closed35.8%
6GLM-5.2Z.AI · Open weight33.7%
7GPT-5.4OpenAI · Closed33.3%
8Claude Opus 4.6 (Adaptive)Anthropic · Closed33.0%
9Gemini 3.1 ProGoogle · Closed32.0%
10Kimi K2.6Moonshot AI · Open weight28.5%
11GPT-5.4 miniOpenAI · Closed28.2%
12GPT-5.4 nanoOpenAI · Closed24.9%
13DeepSeek V4 Pro (Max)DeepSeek · Open weight24.3%
14Qwen3.7 PlusAlibaba · Closed22.4%
15Grok 4.3xAI · Closed17.0%
16Qwen3.5 397BAlibaba · Open weight15.3%
17Qwen3.5 397B (Reasoning)Alibaba · Open weight15.3%
18Step 3.7 FlashStepFun · Open weight14.8%
19GLM-5Z.AI · Open weight14.5%
20Gemini 3.1 Flash-LiteGoogle · Closed12.2%
21Kimi K2.5Moonshot AI · Open weight11.5%
22Kimi K2.5 (Reasoning)Moonshot AI · Closed11.5%
23MiniMax M2.7MiniMax · Open weight10.6%
24GPT-OSS 120BOpenAI · Open weight3.1%
25MiMo-V2.5-ProXiaomi · Closed2.4%
26Nemotron 3 Super 120B A12BNVIDIA · Open weight1.8%
27GPT-OSS 20BOpenAI · Open weight0.7%