Benchmarks · knowledge

AA-Omniscience v1.1 Science, Engineering & Math

AA-Omniscience v1.1 Index (-100..100) on the Science, Engineering & Math questions only: right answers minus made-up ones, abstaining scores 0.

CategoryVersion

28 models · 76 results · data as of 2026-10-29 · artificialanalysis.ai ↗

All results

76 rows

#	Model	Effort	Score
1	Gemini 4 Carbon	high	45.7
2	Gemini 4 Argon	high	45.3
3	Opus 5.5	max	40.8
4	GPT-6 Astra	high	36.0
5	GPT-6 Astra	xhigh	34.0
6	GPT-6 Astra	medium	33.8
7	Fable 5.1	max	33.5
8	GPT-6 Astra	max	33.5
9	GPT-6 Astra	low	33.2
10	Opus 5.5	high	32.7
11	Grok 4.7	xhigh	32.5
12	Fable 5.1	xhigh	32.3
13	GPT-6.1 Sol	medium	32.3
14	Opus 5.5	xhigh	32.2
15	GPT-6.1 Sol	xhigh	32.0
16	Grok 4.7	high	31.8
17	Fable 5.1	high	31.3
18	GPT-6.1 Sol	high	30.8
19	GPT-6.1 Sol	max	30.7
20	Opus 5.5	low	29.5
21	Fable 5.1	medium	27.7
22	Muse Spark 1.3	xhigh	27.7
23	GPT-6.1 Sol	low	27.5
24	Muse Spark 1.3	max	26.3
25	Sonnet 5.5	max	25.8
26	Fable 5.1	low	25.5
27	Grok 4.7	low	25.2
28	Opus 5.5	medium	25.0
29	GPT-6.1 Terra	max	24.0
30	Gemini 3.8 Flash	high	22.5