omniscience Accuracy

114 models evaluated

#	Model	Provider	Type	Score
1	GPT-5.5	OpenAI	Closed	56.9%
2	Gemini 3 Pro	Google	Closed	55.9%
3	Gemini 3.1 Pro	Google	Closed	55.3%
4	Gemini 3.5 Flash	Google	Closed	51.9%
5	GPT-5.3 Codex	OpenAI	Closed	51.8%
6	GPT-5.4	OpenAI	Closed	50.0%
7	Claude Opus 4.6 (Adaptive)	Anthropic	Closed	46.4%
8	Claude Opus 4.7 (Adaptive)	Anthropic	Closed	45.8%
9	Claude Opus 4.5 Thinking	Anthropic	Closed	45.7%
10	Gemini 3 Flash	Google	Closed	45.5%
11	Claude Opus 4.6	Anthropic	Closed	45.2%
12	Muse Spark	Meta	Closed	44.6%
13	GPT-5.2	OpenAI	Closed	43.8%
14	Claude Opus 4.7	Anthropic	Closed	43.5%
15	DeepSeek V4 Pro (Max)	DeepSeek	Open	43.3%
16	DeepSeek V4 Pro (High)	DeepSeek	Open	41.8%
17	Grok 4	xAI	Closed	41.4%
18	GPT-5 (high)	OpenAI	Closed	40.7%
19	GPT-5.2-Codex	OpenAI	Closed	40.7%
20	Claude Opus 4.5	Anthropic	Closed	40.7%
21	GPT-5.1-Codex-Max	OpenAI	Closed	39.2%
22	GPT-5.1-Codex	OpenAI	Closed	39.2%
23	Gemini 2.5 Pro	Google	Closed	39.0%
24	GPT-5 (medium)	OpenAI	Closed	38.9%
25	o3	OpenAI	Closed	38.4%
26	Claude Sonnet 4.6	Anthropic	Closed	38.0%
27	Qwen 3.6 Max (preview)	Alibaba	Closed	37.7%
28	GPT-5.1	OpenAI	Closed	37.6%
29	GPT-5.4 mini	OpenAI	Closed	37.5%
30	DeepSeek V4 Flash (Max)	DeepSeek	Open	37.2%
31	Gemini 3.1 Flash-Lite	Google	Closed	36.4%
32	DeepSeek V4 Flash (High)	DeepSeek	Open	35.5%
33	o1	OpenAI	Closed	34.7%
34	Grok 4.3	xAI	Closed	34.6%
35	Kimi K2.5 (Reasoning)	Moonshot AI	Closed	34.3%
36	Kimi K2.5	Moonshot AI	Open	34.3%
37	Kimi K2.6	Moonshot AI	Open	32.8%
38	Qwen3.5 397B (Reasoning)	Alibaba	Open	31.4%
39	DeepSeek-R1	DeepSeek	Open	31.0%
40	Qwen3.7 Max	Alibaba	Closed	30.1%
41	GLM-4.7	Z.AI	Open	29.3%
42	GLM-5V-Turbo	Z.AI	Closed	29.1%
43	GLM-5-Turbo	Z.AI	Closed	29.0%
44	DeepSeek V3.1 (Reasoning)	DeepSeek	Open	28.8%
45	Hy3 Preview	Tencent	Open	28.0%
46	GLM-5	Z.AI	Open	26.9%
47	Kimi K2	Moonshot AI	Closed	26.8%
48	MiMo-V2-Pro	Xiaomi	Closed	26.8%
49	Gemini 2.5 Flash	Google	Closed	26.5%
50	Qwen3.6 Plus	Alibaba	Closed	26.2%
51	MiniMax M2.7	MiniMax	Open	26.1%
52	DeepSeek V3	DeepSeek	Open	25.4%
53	GPT-5.4 nano	OpenAI	Closed	25.4%
54	Grok 4.1 Fast (Reasoning)	xAI	Closed	25.3%
55	Mistral Medium 3.5 128B	Mistral	Open	25.1%
56	Qwen3.5-122B-A10B	Alibaba	Open	24.7%
57	Qwen3 Max	Alibaba	Closed	24.4%
58	Qwen3.5 397B	Alibaba	Open	24.3%
59	Llama 4 Maverick	Meta	Open	24.3%
60	GLM-5.1	Z.AI	Open	24.2%
61	DeepSeek V3.2	DeepSeek	Open	24.2%
62	GPT-4.1	OpenAI	Closed	24.2%
63	Mistral Large 3	Mistral	Closed	24.1%
64	Grok Code Fast 1	xAI	Closed	23.8%
65	DeepSeek V3.1	DeepSeek	Open	23.1%
66	Trinity-Large-Preview	Arcee AI	Open	22.8%
67	Trinity-Large-Thinking	Arcee AI	Open	22.8%
68	MiMo-V2.5-Pro	Xiaomi	Closed	22.6%
69	Grok 4 Fast (Reasoning)	xAI	Closed	22.6%
70	Claude 4 Sonnet	Anthropic	Closed	22.4%
71	Llama 3.1 405B	Meta	Open	22.3%
72	Mistral Small 4 (Reasoning)	Mistral	Open	22.1%
73	Mistral Small 4	Mistral	Open	22.1%
74	GPT-OSS 120B	OpenAI	Open	21.5%
75	Qwen3.5-27B	Alibaba	Open	21.0%
76	GLM-4.6	Z.AI	Open	20.8%
77	Qwen3.5-35B-A3B	Alibaba	Open	20.5%
78	Mistral Large 2	Mistral	Closed	20.1%
79	Nemotron Ultra 253B	NVIDIA	Open	19.9%
80	Gemma 4 31B	Google	Open	19.9%
81	GPT-4o	OpenAI	Closed	19.7%
82	Qwen3.6-27B	Alibaba	Open	19.2%
83	Qwen3.6-35B-A3B	Alibaba	Open	18.9%
84	MiMo-V2-Omni	Xiaomi	Closed	18.7%
85	Mistral Medium 3	Mistral	Closed	18.3%
86	Gemma 4 26B A4B	Google	Open	18.2%
87	Sarvam 105B	Sarvam	Open	17.6%
88	GPT-4.1 mini	OpenAI	Closed	17.5%
89	Claude 3 Haiku	Anthropic	Closed	17.2%
90	Grok 4.1 Fast	xAI	Closed	17.0%
91	Nova Pro	Amazon	Closed	17.0%
92	K-Exaone	LG AI Research	Closed	16.5%
93	Solar Pro 2	Upstage	Closed	15.6%
94	GLM-4.5-Air	Z.AI	Closed	15.5%
95	GPT-OSS 20B	OpenAI	Open	15.5%
96	Ling 2.6 Flash	InclusionAI	Open	15.4%
97	MiMo-V2-Flash	Xiaomi	Open	15.2%
98	Nemotron 3 Nano Omni 30B A3B	NVIDIA	Open	14.8%
99	Llama 4 Scout	Meta	Open	14.6%
100	GPT-4.1 nano	OpenAI	Closed	13.3%
101	Phi-4	Microsoft	Open	13.2%
102	Sarvam 30B	Sarvam	Open	12.7%
103	Gemma 3 27B	Google	Open	12.5%
104	Nemotron 3 Nano 30B	NVIDIA	Open	11.4%
105	Exaone 4.0 32B	LG AI Research	Open	10.4%
106	Command A+	Cohere	Open	8.9%
107	LFM2.5-8B-A1B	LiquidAI	Open	8.7%
108	Gemma 4 E4B	Google	Open	8.6%
109	Gemma 4 E2B	Google	Open	6.7%
110	Granite-4.0-1B	IBM	Open	6.1%
111	Granite-4.0-H-1B	IBM	Open	5.3%
112	Exaone 4.0 1.2B	LG AI Research	Open	4.7%
113	Granite-4.0-H-350M	IBM	Open	3.7%
114	Granite-4.0-350M	IBM	Open	3.2%