AA-AnalystAgent · Open Analysis

#	Model	Effort	Score	Cost / task	Harness
1	Gemini 3.7 Flash	high	60.0%	–	–
2	Sonnet 5.5	max	57.5%	–	–
3	Fable 5.1	max	57.5%	–	–
4	Opus 5.5	max	56.3%	–	–
5	Grok 4.7	high	55.0%	–	–
6	Opus 5	max	53.8%	–	–
7	Haiku 5.5	max	53.8%	–	–
8	GPT-6 Astra	max	51.2%	–	–
9	GPT-6.1 Sol	max	50.0%	–	–
10	GPT-5.5	xhigh	50.0%	–	–
11	Gemini 4	high	48.8%	–	–
12	Fable 5	max	48.8%	–	–
13	GPT-5.6 Sol	max	47.5%	–	–
14	Gemini 3.8 Flash	high	46.3%	–	–
15	Sonnet 5	max	46.3%	–	–
16	Qwen 3.8 Max	max	45.0%	–	–
17	Opus 4.8	max	45.0%	–	–
18	Gemini 3.5 Flash	high	45.0%	–	–
19	Opus 4.7	max	43.8%	–	–
20	GPT-6 Luna	max	43.8%	–	–
21	Grok 4.6	high	41.3%	–	–
22	Gemini 3.1 Pro	–	41.3%	–	–
23	GLM 5.3	max	38.8%	–	–
24	Kimi K3	max	38.8%	–	–
25	Muse Spark 1.3	max	37.5%	–	–
26	MiMo V2.6 Pro	–	37.5%	–	–
27	Step 5 Preview	–	35.0%	–	–
28	DeepSeek V4.1 Flash	max	35.0%	–	–
29	Grok 4.5	high	35.0%	–	–
30	Inkling Small	–	27.5%	–	–
31	GLM 5.3 Flash	max	25.0%	–	–
32	DeepSeek V4 Flash 0420	max	25.0%	–	–
33	Inkling	xhigh	23.8%	–	–
34	Sonnet 4.6	max	20.0%	–	–
35	MiMo V2.5 Pro	–	20.0%	–	–
36	Muse Glimmer (High)	high	20.0%	–	–
37	DeepSeek V4 Pro 0424	max	18.8%	–	–
38	Qwen 3.7 Max	max	18.8%	–	–
39	Ling 3.0 Flash Fin	–	16.3%	–	–
40	GPT-5.4 Mini	xhigh	15.0%	–	–
41	Haiku 4.5	–	15.0%	–	–
42	Mistral Medium 3.5	high	12.5%	–	–
43	MiniMax M2.7	–	11.3%	–	–
44	MiniMax M3	–	10.0%	–	–
45	Grok 4.3	high	8.8%	–	–
46	Gemini 3.1 Flash-Lite	high	8.8%	–	–
47	Nemotron 3 Ultra	max	6.3%	–	–
48	Mistral Small 4	–	1.3%	–	–