Benchmarks · knowledge

AA-Omniscience Software Engineering

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

CategoryVersion

136 models · 267 results · data as of 2026-10-29 · artificialanalysis.ai ↗

All results

267 rows

#	Model	Effort	Score
1	Opus 5.5	max	88.8
2	Fable 5.1	max	86.2
3	Opus 5.5	xhigh	84.9
4	GPT-6 Astra	xhigh	84.2
5	Opus 5.5	medium	83.4
6	Opus 5.5	high	83.3
7	Fable 5.1	xhigh	83.0
8	GPT-6 Astra	medium	83.0
9	GPT-6 Astra	max	82.8
10	GPT-6 Astra	high	82.6
11	GPT-6.1 Sol	high	82.2
12	GPT-6.1 Sol	max	81.6
13	Opus 5.5	low	81.3
14	GPT-6 Astra	low	80.9
15	GPT-6.1 Sol	medium	80.7
16	GPT-6.1 Sol	xhigh	80.7
17	GPT-6.1 Sol	low	80.1
18	Fable 5.1	high	79.9
19	Sonnet 5.5	max	78.3
20	Fable 5.1	medium	77.3
21	Opus 5	max	74.3
22	Sonnet 5.5	xhigh	74.0
23	GPT-5.5	xhigh	72.6
24	GPT-5.5	high	71.9
25	Fable 5.1	low	71.8
26	GPT-6.1 Terra	high	71.7
27	Gemini 4 Argon	high	71.4
28	GPT-5.6 Sol	max	71.4
29	GPT-5.5	medium	71.3
30	Opus 5	xhigh	71.2