AA-Omniscience Accuracy

Correct answers divided by all questions. AA v4.3 gives this component 10%.

365models

–Open Index weight

–last updated

Artificial Analysispublisher

#	Model	Effort	Score	Cost / task	Harness
1	Fable 5.1	max	67.2%	$0.268	–
2	Opus 5.5	max	66.2%	$0.161	–
3	Fable 5	max	65.3%	$0.070	–
4	GPT-6 Astra	max	62.6%	$0.097	–
5	GPT-6.1 Sol	max	62.1%	$0.017	–
6	Opus 5	max	60.9%	$0.026	–
7	GPT-5.6 Sol	max	59.4%	$0.085	–
8	GPT-5.5	xhigh	58.0%	$0.152	–
9	Gemini 3 Pro	high	55.8%	–	–
10	Gemini 3.7 Flash	high	55.3%	$0.0039	–
11	Gemini 3.1 Pro	–	54.9%	$0.018	–
12	Gemini 3.8 Flash	high	54.6%	$0.013	–
13	GPT-6 Sol	max	54.5%	$0.027	–
14	Sonnet 5.5	max	53.9%	$0.147	–
15	Gemini 3 Flash	–	53.4%	–	–
16	GPT-5.3 Codex	xhigh	52.9%	–	–
17	Muse Spark 1.1	xhigh	52.0%	$0.014	–
18	Grok 4.5	high	51.5%	$0.0065	–
19	Grok Build 0.1	–	51.5%	–	–
20	Gemini 3.5 Flash	high	51.4%	$0.014	–
21	GPT-5.4	xhigh	50.8%	–	–
22	Gemini 3.6 Flash	high	50.0%	$0.0052	–
23	Gemini 4	high	49.9%	$0.0096	–
24	Grok 4.7	low	49.7%	$0.011	–
25	Muse Spark	–	49.6%	–	–
26	JT-4.1 Flash 236B A21B (Reasoning)	–	49.4%	–	–
27	DeepSeek V4 Pro 0813	max	49.1%	$0.013	–
28	Opus 4.7	max	48.9%	–	–
29	Opus 4.8	max	48.8%	$0.025	–
30	Grok 4.6	high	48.2%	$0.0087	–
31	Kimi K3	max	47.6%	$0.133	–
32	Opus 4.6	max	47.0%	–	–
33	GPT-5.6 Terra	max	46.8%	$0.070	–
34	Opus 4.5	–	46.6%	–	–
35	DeepSeek V4.1 Flash	max	46.4%	$0.0098	–
36	GPT-5.5 Instant	–	46.3%	–	–
37	Muse Spark 1.2	xhigh	45.4%	$0.010	–
38	GPT-5.2	xhigh	44.3%	–	–
39	GPT-6 Luna	xhigh	44.2%	$0.0010	–
40	Muse Spark 1.3	max	43.6%	$0.023	–
41	Muse Wonder Full (Max)	max	43.2%	–	–
42	DeepSeek V4 Pro 0424	max	43.0%	$0.0056	–
43	GPT-5.6 Luna	max	42.7%	$0.011	–
44	Inkling	xhigh	41.5%	–	–
45	Step 5 Preview	–	41.5%	$0.015	–
46	Muse Wonder	max	41.1%	$0.050	–
47	GPT-5.2 Codex	xhigh	41.1%	–	–
48	Sonnet 4.6	max	40.9%	$0.044	–
49	Grok 4	–	40.5%	–	–
50	DeepSeek V4 Flash 0731	max	40.4%	$0.020	–
51	GPT-5	high	40.3%	–	–
52	Sonnet 5	max	40.1%	$0.029	–
53	GPT-5.1 Codex	high	39.9%	–	–
54	Kimi K2.7 Code	–	39.6%	$0.021	–
55	Muse Spark Open (Xhigh)	xhigh	39.3%	$0.0080	–
56	Gemini 2.5 Pro	–	39.1%	$0.017	–
57	GPT-5 Codex	high	38.7%	–	–
58	DeepSeek V4 Flash	max	38.6%	$0.0058	–
59	o3	–	38.6%	–	–
60	Qwen 3.6 Max	max	37.9%	–	–
61	DeepSeek V3.2	–	37.8%	–	–
62	GPT-5.1	high	37.7%	–	–
63	GPT-5.4 Mini	xhigh	37.5%	$0.057	–
64	DeepSeek V4 Flash 0420	max	36.8%	$0.0016	–
65	Haiku 5.5	max	36.4%	$0.0063	–
66	Gemini 3.1 Flash-Lite	high	36.3%	$0.0016	–
67	Kimi K2.5	–	35.2%	–	–
68	MiMo V2.6 Pro	–	34.8%	$0.0040	–
69	Nex-N2-Pro	–	34.8%	–	–
70	Grok 4.3	high	34.8%	$0.0075	–
71	o1	–	34.5%	–	–
72	GLM 5.3	low	33.9%	$0.0008	–
73	Inkling Small	–	33.2%	$0.0060	–
74	Beam 1.0	–	33.1%	–	–
75	Sonnet 4.5	–	32.9%	$0.013	–
76	Kimi K2.6	–	32.6%	$0.026	–
77	Hy3	–	32.0%	$0.0035	–
78	Qwen 3.8 Max	max	31.9%	$0.029	–
79	Apodex 1.1	–	31.7%	$0.015	–
80	Qwen 3.8 2.4T-A95B	–	31.3%	$0.028	–
81	Qwen 3.7 Max	max	31.1%	$0.014	–
82	Kimi K2	–	30.9%	–	–
83	Qwen 3.5 397B-A17B	–	30.8%	$0.015	–
84	DeepSeek R1	–	30.5%	–	–
85	Cogito v2.1 (Reasoning)	–	30.2%	–	–
86	Llama 3.1 405b	–	30.1%	–	–
87	Motif 3	–	30.1%	–	–
88	Qwen 3 Max	max	29.8%	–	–
89	LongCat 2.0	–	29.6%	$0.024	–
90	Gemini 3.5 Flash-Lite	high	29.5%	$0.0031	–
91	GLM 5 V	–	29.3%	–	–
92	GLM 4.7	–	29.3%	–	–
93	Ling 3.1 Flash	–	29.1%	$0.0040	–
94	DeepSeek V3.1	–	29.0%	–	–
95	Grok 3	none	28.9%	–	–
96	Grok 4.20	–	28.9%	–	–
97	GLM 5	–	28.4%	–	–
98	Gemini 2.5 Flash	–	28.3%	–	–
99	Sonnet 3.7	none	28.2%	–	–
100	GPT-4.1	none	27.8%	–	–
101	GLM 5.3 Flash	max	27.5%	$0.0006	–
102	MiMo V2.5 Pro	none	27.2%	–	–
103	MiMo V2.6 Flash	–	27.0%	$0.0033	–
104	Muse Glimmer (High)	high	27.0%	$0.0025	–
105	GLM 4.6	–	26.9%	–	–
106	MiniMax M2.7	–	26.8%	–	–
107	MiMo V2 Pro	–	26.6%	–	–
108	Qwen 3.6 Plus	–	26.4%	–	–
109	MiniMax M2.5	–	26.2%	–	–
110	Mistral Large 4	–	25.8%	$0.017	–
111	Step 3.7 Flash	high	25.8%	–	–
112	Ring-2.6-1T	xhigh	25.7%	$0.0049	–
113	GPT-5.4 Nano	xhigh	25.7%	$0.0099	–
114	DeepSeek V3	none	25.4%	$0.0001	–
115	GLM 5.1	none	25.2%	–	–
116	GLM 4.5	–	25.1%	–	–
117	Grok 4.1 Fast	–	25.1%	–	–
118	GPT-5 Mini	high	25.0%	$0.0056	–
119	Mistral Large 3	none	24.9%	$0.0001	–
120	Llama 4 Maverick	none	24.9%	$0.0001	–
121	Doubao Seed Code	–	24.9%	–	–
122	MiMo V2 Flash	–	24.8%	–	–
123	o4 mini	high	24.8%	–	–
124	Mistral Medium 3.5	high	24.7%	$0.014	–
125	Step 3.5 Flash 2603	–	24.6%	–	–
126	Qwen 3.8 Flash	–	24.5%	$0.0021	–
127	Qwen 3.5 122B-A10B	–	24.4%	$0.014	–
128	GLM 5.2	max	24.3%	$0.0086	–
129	Nemotron 3 Super 120B A12B (Reasoning)	–	24.3%	$0.0021	–
130	JT-4.1 Flash 236B A21B (Non-reasoning)	none	24.3%	–	–
131	JT-35B-Flash	none	24.2%	–	–
132	GPT-4o	none	23.7%	–	–
133	Step 3.5 Flash	–	23.6%	–	–
134	Grok Code Fast 1	–	23.5%	–	–
135	Llama 3.1 70b	–	23.5%	–	–
136	Gemini 2.0 Flash	none	23.5%	–	–
137	Nemotron 3.5 Super VL	–	23.3%	$0.0033	–
138	MiniMax M2	–	23.2%	–	–
139	GPT-5.1 Codex Mini	high	23.2%	–	–
140	Qwen 3 235B-A22B	–	22.8%	–	–
141	Grok 4 Fast	–	22.8%	–	–
142	Sonnet 4	–	22.7%	–	–
143	Mercury 2.5	–	22.7%	$0.0015	–
144	Nemotron 3 Ultra	max	22.6%	$0.0031	–
145	Qwen 3.7 Plus	–	22.5%	$0.0029	–
146	Trinity Large Thinking	–	22.5%	$0.0050	–
147	MiniMax M1	–	22.5%	–	–
148	ERNIE 5.0 Thinking Preview	–	22.1%	–	–
149	Nova 2.0 Pro Preview (Low)	low	22.0%	–	–
150	Nova 2.0 Pro Preview (Medium)	medium	22.0%	–	–
151	Ling-2.6-1T	none	21.9%	–	–
152	gpt-oss-120b	high	21.8%	$0.0011	–
153	Mistral Small 4	–	21.7%	$0.0004	–
154	KAT Coder Pro V2	none	21.5%	–	–
155	o3 mini	high	21.3%	–	–
156	Nova Premier	none	21.3%	–	–
157	Mercury 2	high	21.2%	$0.0035	–
158	MiniMax M2.1	–	21.2%	–	–
159	Motif 3 (Beta)	–	21.0%	–	–
160	GPT-4	none	21.0%	–	–
161	Qwen 3	–	20.8%	–	–
162	GLM 4.5 V	–	20.8%	–	–
163	Devstral 2	none	20.8%	–	–
164	Magistral Medium 1.2	medium	20.8%	–	–
165	Mistral Medium 3.1	none	20.8%	–	–
166	Qwen 3.5 27B	–	20.7%	–	–
167	Jamba 1.7 Large	none	20.6%	–	–
168	Ring-1T	–	20.5%	–	–
169	Ling-1T	none	20.3%	–	–
170	Magistral Medium 1	medium	20.3%	–	–
171	GPT-4.1 Mini	none	20.3%	–	–
172	Qwen 3.5 35B-A3B	–	20.1%	–	–
173	Llama 3.1	max	20.1%	–	–
174	Gemma 4 31B	–	20.0%	$0.0000	–
175	Solar Pro 2 (Reasoning)	high	20.0%	–	–
176	Mistral Large 2	none	19.9%	–	–
177	Qwen 3.6 27B	–	19.6%	$0.034	–
178	INTELLECT-3 (Based on GLM-4.5-Air)	–	19.6%	–	–
179	Solar Pro 4	–	19.4%	$0.0029	–
180	MiMo V2	–	19.3%	–	–
181	Gemma 4 26B	–	19.1%	–	–
182	GPT-5 Nano	high	19.1%	–	–
183	ERNIE 4.5 300B A47B	none	19.1%	–	–
184	Nova 2.0 Lite (High)	high	19.0%	–	–
185	Llama 3.3 70b	none	18.9%	–	–
186	North Mini Code	–	18.9%	$0.0000	–
187	Devstral Medium	none	18.9%	–	–
188	Solar Open2 250B	–	18.8%	–	–
189	Qwen 3.6 35B-A3B	–	18.8%	$0.0049	–
190	A.X-K2	–	18.6%	–	–
191	Nova 2.0 Omni (Low)	low	18.6%	–	–
192	Solar Open 100B (Reasoning)	high	18.5%	–	–
193	K2-V2 (High)	high	18.5%	–	–
194	Solar Pro 3	high	18.5%	$0.0013	–
195	Solar Mini 4	–	18.4%	$0.0069	–
196	Qwen 3.8 27B	medium	18.4%	$0.0087	–
197	Mistral Medium 3	none	18.3%	–	–
198	Nemotron Cascade 2 30B A3B	–	18.3%	–	–
199	Seed-OSS-36B-Instruct	–	18.2%	–	–
200	K2 Horizon 375B A23B	–	18.2%	–	–
201	Ling 3.0 Flash	–	18.2%	–	–
202	Nova 2.0 Lite (Medium)	medium	18.2%	–	–
203	Haiku 4.5	–	18.0%	$0.0045	–
204	K2 Think V2	high	18.0%	–	–
205	Magistral Small 1	–	18.0%	–	–
206	Gemini 2.5 Flash Lite	–	18.0%	–	–
207	K2-V2 (Medium)	medium	17.9%	–	–
208	Mi:dm K 2.5 Pro	–	17.9%	–	–
209	Ling 3.0 Flash Fin	–	17.9%	–	–
210	Llama 3 70b	none	17.7%	–	–
211	KAT-Coder-Pro V1	none	17.7%	–	–
212	Nova 2.0 Omni (Medium)	medium	17.7%	–	–
213	Nova 2.0 Lite (Low)	low	17.6%	–	–
214	Haiku 3	none	17.6%	–	–
215	Sarvam 105B (High)	high	17.6%	–	–
216	Qwen 2.5 72B	none	17.5%	–	–
217	Qwen 3.5	none	17.4%	–	–
218	Qwen 3 32B	–	17.4%	–	–
219	GLM 4.6 V	none	17.4%	–	–
220	NVIDIA Nemotron 3 Nano 30B A3B (Reasoning)	–	17.3%	$0.0010	–
221	Llama Nemotron Super 49B v1.5 (Reasoning)	–	17.2%	–	–
222	Llama 3.3	–	17.1%	–	–
223	Nemotron 3.5 Bolt 74B-A9B	–	17.1%	–	–
224	Apriel-v1.6-15B-Thinker	–	17.0%	–	–
225	Nova 2.0 Pro Preview (Non-reasoning)	none	16.9%	–	–
226	Nova Pro	none	16.9%	–	–
227	Agnes 2.5 Pro Beta	–	16.8%	–	–
228	MiMo V2.5	–	16.8%	$0.0005	–
229	MiniMax M3	–	16.7%	$0.0035	–
230	DiffusionGemma 26B A4B	–	16.6%	–	–
231	Ring-flash-2.0	–	16.6%	–	–
232	Apriel-v1.5-15B-Thinker	–	16.4%	–	–
233	Qwen 3.5 9B	–	16.4%	$0.0028	–
234	Mi:dm K 2.5 Pro Preview	–	16.4%	–	–
235	Qwen 3 Coder	none	16.4%	–	–
236	Command A	none	16.4%	–	–
237	K-EXAONE (Reasoning)	–	16.4%	–	–
238	K2-V2 (Low)	low	16.3%	–	–
239	Qwen 3 30B-A3B	–	16.3%	$0.0015	–
240	GLM 4.5 Air	–	16.3%	–	–
241	EXAONE 4.5 33B (Reasoning)	–	16.2%	–	–
242	GLM 4.7 Flash	–	16.2%	–	–
243	Solar Pro 2 (Non-reasoning)	none	16.1%	–	–
244	gpt-oss-20b	high	16.0%	$0.0004	–
245	Devstral Small	none	15.9%	–	–
246	Devstral Small 2	none	15.8%	–	–
247	Gemma 4 12B	–	15.6%	–	–
248	Ling 2.6 Flash	none	15.6%	–	–
249	Quasar 438B (Max, Based on GLM-5.2)	max	15.5%	$0.018	–
250	HyperCLOVA X SEED Think (32B)	–	15.4%	–	–
251	Qwen 3 14B	–	15.3%	–	–
252	Sarvam M (Reasoning, Based on Mistral Small 3.1)	high	15.2%	–	–
253	HyperNova 60B 2605 (High, Based on gpt-oss-120b)	high	15.2%	–	–
254	Nemotron 3 Nano Omni 30B A3B Reasoning	–	15.2%	–	–
255	Motif-2-12.7B-Reasoning	–	15.2%	–	–
256	Llama 4 Scout	none	15.2%	$0.0000	–
257	Grok 3 Mini	high	15.1%	–	–
258	Qwen 3.5 4B	–	15.1%	–	–
259	Mistral Small 3.1	none	15.1%	$0.0000	–
260	K2 Horizon MoVA 36B A4B	–	15.0%	–	–
261	G9v3-39A5B	–	14.9%	–	–
262	Olmo 3 32B Think	–	14.9%	–	–
263	Mistral Small 3.2	none	14.8%	$0.0000	–
264	NVIDIA Nemotron Nano 12B v2 VL (Reasoning)	–	14.6%	–	–
265	Falcon-H1R-7B	–	14.5%	–	–
266	Nemotron 3.5 Lightning	–	14.4%	$0.0006	–
267	Ling 3.0 Flash	–	14.3%	–	–
268	Granite 4.0 H Small	none	14.3%	–	–
269	JT-MINI	none	14.2%	–	–
270	Ling-flash-2.0	none	14.1%	–	–
271	Phi-4	none	14.1%	–	–
272	Nova 2.0 Lite (Non-reasoning)	none	14.0%	–	–
273	Granite 4.1 30B	none	14.0%	–	–
274	EXAONE 4.0 32B (Reasoning)	–	13.9%	–	–
275	Olmo 3.1 32B Think	–	13.8%	–	–
276	GPT-4.1 Nano	none	13.7%	–	–
277	Reka Flash 3	–	13.7%	–	–
278	K-EXAONE (Non-reasoning)	none	13.7%	–	–
279	Qwen 3 8B	–	13.6%	–	–
280	Ministral 3 14B	none	13.6%	$0.0000	–
281	Magistral Small 1.2	–	13.5%	–	–
282	LongCat Flash Lite	none	13.3%	–	–
283	Step3 VL 10B	–	13.3%	–	–
284	Nova 2.0 Omni (Non-reasoning)	none	13.2%	–	–
285	Haiku 3.5	none	13.2%	–	–
286	K-EXAONE 2.0 0803	–	13.1%	–	–
287	Ministral 3 8B	none	13.0%	$0.0000	–
288	Gemma 3 27B	none	13.0%	$0.0000	–
289	K2 Horizon 7B	–	12.8%	–	–
290	Sarvam 30B (High)	high	12.6%	–	–
291	Llama Nemotron Super 49B v1.5 (Non-reasoning)	none	12.6%	–	–
292	Qwen 3 4B	–	12.5%	–	–
293	Tri-21B-Think	–	12.3%	–	–
294	Granite 4.1 8B	none	12.3%	–	–
295	Olmo 3.1 32B Instruct	none	12.0%	–	–
296	NVIDIA Nemotron Nano 12B v2 VL (Non-reasoning)	none	11.9%	–	–
297	NVIDIA Nemotron Nano 9B V2 (Reasoning)	–	11.8%	–	–
298	Apertus 70B Instruct	none	11.6%	–	–
299	Molmo2-8B	none	11.5%	–	–
300	NVIDIA Nemotron 3 Nano 30B A3B (Non-reasoning)	none	11.4%	–	–
301	Jamba 1.7 Mini	none	11.3%	–	–
302	Granite 4.2 8B	–	11.2%	$0.0013	–
303	Olmo 3 7B Think	–	11.1%	–	–
304	Celeris-1	none	11.0%	$0.0000	–
305	K2 Horizon 3.7B	–	10.8%	–	–
306	EXAONE 4.0 32B (Non-reasoning)	none	10.6%	–	–
307	Llama 3.2 11b	none	10.6%	–	–
308	Llama 3 8b	none	10.4%	–	–
309	Gemma 3 12B	none	10.3%	$0.0000	–
310	Tri-21B-think Preview	–	10.1%	–	–
311	Granite 4.2 30B	–	10.1%	$0.0013	–
312	Nova Micro	none	10.0%	–	–
313	Nanbeige4.1-3B	–	9.9%	–	–
314	NVIDIA Nemotron Nano 9B V2 (Non-reasoning)	none	9.8%	–	–
315	Nova Lite	none	9.8%	–	–
316	Granite 3.3 8B (Non-reasoning)	none	9.7%	–	–
317	Apertus 8B Instruct	none	9.7%	–	–
318	NVIDIA Nemotron 3 Nano 4B	–	9.6%	–	–
319	Phi-4 Mini Instruct	none	9.5%	–	–
320	LFM2.5-8B-A1B	–	9.3%	–	–
321	Granite 4.0 Micro	none	9.3%	–	–
322	Granite 4.1 3B	none	9.3%	–	–
323	Granite 4.2 3B	–	9.2%	$0.0006	–
324	Mistral	none	9.0%	–	–
325	Ministral 3 3B	none	9.0%	$0.0000	–
326	Ling-mini-2.0	none	8.9%	–	–
327	Qwen 3 1.7B	–	8.9%	–	–
328	Gemma 4 E4B (Reasoning)	–	8.6%	$0.0000	–
329	Ling 3.0 Tiny	–	8.5%	$0.0000	–
330	Llama 3.1 8b	none	8.5%	–	–
331	MiniCPM5-2B	–	8.4%	–	–
332	Gemma 4 E4B (Non-reasoning)	none	8.3%	–	–
333	LFM2 8B A1B	none	8.2%	–	–
334	Gemma 4 E2B (Non-reasoning)	none	8.2%	–	–
335	Qwen 3.5 2B	–	8.1%	–	–
336	Gemma 3n E4B Instruct	none	8.0%	–	–
337	Gemma 3 4B	none	7.7%	–	–
338	LFM2.5-1.2B-Thinking	–	7.4%	–	–
339	Olmo 3 7B Instruct	none	7.4%	–	–
340	LFM2.5-1.2B-Instruct	none	7.0%	–	–
341	Llama 3.2 1b	none	7.0%	–	–
342	K2 Horizon 0.9B	–	6.9%	–	–
343	Exaone 4.0 1.2B (Reasoning)	–	6.8%	–	–
344	Gemma 3n E2B Instruct	none	6.8%	–	–
345	MiniCPM-V 4.6 1.3B	none	6.7%	–	–
346	Jamba Reasoning 3B	–	6.7%	–	–
347	Gemma 4 E2B (Reasoning)	–	6.6%	–	–
348	G9v3-3B	–	6.5%	–	–
349	LFM2 24B A2B	none	6.5%	–	–
350	Qwen 3 0.6B	–	6.3%	–	–
351	Granite 4.0 1B	none	6.2%	–	–
352	Tiny Aya Global	none	6.1%	–	–
353	LFM2.5-VL-1.6B	none	5.8%	–	–
354	LFM2 2.6B	none	5.4%	–	–
355	Granite 4.0 H 1B	none	5.2%	–	–
356	Exaone 4.0 1.2B (Non-reasoning)	none	5.0%	–	–
357	Qwen 3.5 0.8B	none	5.0%	–	–
358	LFM2 1.2B	none	4.9%	–	–
359	LFM2.5-2.6B	–	4.3%	–	–
360	Granite 4.0 350M	none	3.9%	–	–
361	Granite 4.0 H 350M	none	3.8%	–	–
362	Gemma 3 1B	none	3.8%	–	–
363	MiniCPM5-1B (Reasoning)	–	2.0%	–	–
364	Gemma 3 270M	none	0.9%	–	–
365	MiniCPM5-1B (Non-reasoning)	none	0.1%	–	–