aa Hle

122 models evaluated

#	Model	Provider	Type	Score
1	Gemini 3.1 Pro	Google	Closed	44.7%
2	GPT-5.5	OpenAI	Closed	44.3%
3	GPT-5.4	OpenAI	Closed	41.6%
4	Gemini 3.5 Flash	Google	Closed	41.0%
5	GPT-5.3 Codex	OpenAI	Closed	39.9%
6	Muse Spark	Meta	Closed	39.9%
7	Claude Opus 4.7 (Adaptive)	Anthropic	Closed	39.6%
8	Qwen3.7 Max	Alibaba	Closed	38.1%
9	Gemini 3 Pro	Google	Closed	37.2%
10	Claude Opus 4.6 (Adaptive)	Anthropic	Closed	36.7%
11	DeepSeek V4 Pro (Max)	DeepSeek	Open	35.9%
12	Kimi K2.6	Moonshot AI	Open	35.9%
13	GPT-5.2	OpenAI	Closed	35.4%
14	Grok 4.3	xAI	Closed	35.0%
15	MiMo-V2.5-Pro	Xiaomi	Closed	33.8%
16	DeepSeek V4 Pro (High)	DeepSeek	Open	33.5%
17	GPT-5.2-Codex	OpenAI	Closed	33.5%
18	DeepSeek V4 Flash (Max)	DeepSeek	Open	32.1%
19	Claude Opus 4.7	Anthropic	Closed	31.2%
20	Kimi K2.5 (Reasoning)	Moonshot AI	Closed	29.4%
21	Kimi K2.5	Moonshot AI	Open	29.4%
22	Qwen 3.6 Max (preview)	Alibaba	Closed	28.9%
23	Claude Opus 4.5 Thinking	Anthropic	Closed	28.4%
24	MiMo-V2-Pro	Xiaomi	Closed	28.3%
25	MiniMax M2.7	MiniMax	Open	28.1%
26	GLM-5.1	Z.AI	Open	28.0%
27	DeepSeek V4 Flash (High)	DeepSeek	Open	27.8%
28	Qwen3.5 397B (Reasoning)	Alibaba	Open	27.3%
29	GLM-5	Z.AI	Open	27.2%
30	GPT-5.4 mini	OpenAI	Closed	26.6%
31	GPT-5.1	OpenAI	Closed	26.5%
32	GPT-5 (high)	OpenAI	Closed	26.5%
33	GPT-5.4 nano	OpenAI	Closed	26.5%
34	Qwen3.6 Plus	Alibaba	Closed	25.7%
35	Hy3 Preview	Tencent	Open	25.5%
36	GLM-5-Turbo	Z.AI	Closed	25.4%
37	GLM-4.7	Z.AI	Open	25.1%
38	Grok 4	xAI	Closed	23.9%
39	GPT-5 (medium)	OpenAI	Closed	23.5%
40	GPT-5.1-Codex-Max	OpenAI	Closed	23.4%
41	Qwen3.5-122B-A10B	Alibaba	Open	23.4%
42	GPT-5.1-Codex	OpenAI	Closed	23.4%
43	Gemma 4 31B	Google	Open	22.7%
44	Qwen3.5-27B	Alibaba	Open	22.2%
45	Qwen3.6-27B	Alibaba	Open	21.6%
46	Gemini 2.5 Pro	Google	Closed	21.1%
47	Qwen3.6-35B-A3B	Alibaba	Open	20.2%
48	o3	OpenAI	Closed	20.0%
49	MiMo-V2-Omni	Xiaomi	Closed	19.9%
50	Qwen3.5-35B-A3B	Alibaba	Open	19.7%
51	Qwen3.5 397B	Alibaba	Open	18.8%
52	Claude Opus 4.6	Anthropic	Closed	18.6%
53	GPT-OSS 120B	OpenAI	Open	18.5%
54	Gemma 4 26B A4B	Google	Open	18.3%
55	Grok 4.1 Fast (Reasoning)	xAI	Closed	17.6%
56	Grok 4 Fast (Reasoning)	xAI	Closed	17.0%
57	Gemini 3.1 Flash-Lite	Google	Closed	16.2%
58	GLM-5V-Turbo	Z.AI	Closed	15.8%
59	DeepSeek-R1	DeepSeek	Open	14.9%
60	Trinity-Large-Preview	Arcee AI	Open	14.7%
61	Trinity-Large-Thinking	Arcee AI	Open	14.7%
62	Gemini 3 Flash	Google	Closed	14.1%
63	Claude Sonnet 4.6	Anthropic	Closed	13.2%
64	K-Exaone	LG AI Research	Closed	13.1%
65	DeepSeek V3.1 (Reasoning)	DeepSeek	Open	13.0%
66	Claude Opus 4.5	Anthropic	Closed	12.9%
67	Mistral Medium 3.5 128B	Mistral	Open	12.8%
68	Claude 4.1 Opus Thinking	Anthropic	Closed	11.9%
69	Command A+	Cohere	Open	11.4%
70	Qwen3 Max	Alibaba	Closed	11.1%
71	DeepSeek V3.2	DeepSeek	Open	10.5%
72	Sarvam 105B	Sarvam	Open	10.1%
73	GPT-OSS 20B	OpenAI	Open	9.8%
74	Mistral Small 4 (Reasoning)	Mistral	Open	9.5%
75	Mistral Small 4	Mistral	Open	9.5%
76	o3-mini	OpenAI	Closed	8.7%
77	Nemotron Ultra 253B	NVIDIA	Open	8.1%
78	MiMo-V2-Flash	Xiaomi	Open	8.0%
79	o1	OpenAI	Closed	7.7%
80	Grok Code Fast 1	xAI	Closed	7.5%
81	Kimi K2	Moonshot AI	Closed	7.0%
82	Sarvam 30B	Sarvam	Open	7.0%
83	GLM-4.5-Air	Z.AI	Closed	6.8%
84	Granite-4.0-H-350M	IBM	Open	6.4%
85	DeepSeek V3.1	DeepSeek	Open	6.3%
86	Ling 2.6 Flash	InclusionAI	Open	6.2%
87	Exaone 4.0 1.2B	LG AI Research	Open	5.8%
88	Granite-4.0-350M	IBM	Open	5.7%
89	DeepSeek R1 Distill Qwen 32B	DeepSeek	Open	5.5%
90	Nemotron 3 Nano Omni 30B A3B	NVIDIA	Open	5.3%
91	GLM-4.6	Z.AI	Open	5.2%
92	Gemini 2.5 Flash	Google	Closed	5.1%
93	Granite-4.0-1B	IBM	Open	5.1%
94	Grok 4.1 Fast	xAI	Closed	5.0%
95	Granite-4.0-H-1B	IBM	Open	5.0%
96	Gemini 1.5 Pro	Google	Closed	4.9%
97	Exaone 4.0 32B	LG AI Research	Open	4.9%
98	Llama 4 Maverick	Meta	Open	4.8%
99	Gemma 4 E2B	Google	Open	4.8%
100	Gemma 3 27B	Google	Open	4.7%
101	GPT-4.1	OpenAI	Closed	4.6%
102	GPT-4.1 mini	OpenAI	Closed	4.6%
103	Nemotron 3 Nano 30B	NVIDIA	Open	4.6%
104	Gemini 1.0 Pro	Google	Closed	4.6%
105	Llama 4 Scout	Meta	Open	4.3%
106	Mistral Medium 3	Mistral	Closed	4.3%
107	Llama 3.1 405B	Meta	Open	4.2%
108	Mistral Large 3	Mistral	Closed	4.1%
109	Phi-4	Microsoft	Open	4.1%
110	Claude 4 Sonnet	Anthropic	Closed	4.0%
111	GPT-4o mini	OpenAI	Closed	4.0%
112	Mistral Large 2	Mistral	Closed	4.0%
113	GPT-4.1 nano	OpenAI	Closed	3.9%
114	Claude 3 Haiku	Anthropic	Closed	3.9%
115	Qwen2.5 Coder 32B Instruct	Alibaba	Open	3.8%
116	Solar Pro 2	Upstage	Closed	3.8%
117	Gemma 4 E4B	Google	Open	3.7%
118	DeepSeek V3	DeepSeek	Open	3.6%
119	Nova Pro	Amazon	Closed	3.4%
120	GPT-4o	OpenAI	Closed	3.3%
121	GPT-4 Turbo	OpenAI	Closed	3.3%
122	Claude 3 Opus	Anthropic	Closed	3.1%