Arena Search · Open Analysis

#	Model	Effort	Score	Cost / task	Harness
1	GPT-5.6 Sol	xhigh	1261	–	–
2	Qwen 3.8 Max	max	1261	–	–
3	Opus 4.6	max	1251	–	–
4	GPT-5.5	xhigh	1247	–	–
5	Fable 5.1	high	1246	–	–
6	Opus 5	high	1242	–	–
7	GPT-6 Astra	max	1234	–	–
8	GPT-6.1 Sol	max	1233	–	–
9	Opus 4.7	max	1230	–	–
10	Fable 5	high	1229	–	–
11	Sonnet 4.6	max	1221	–	–
12	Grok 4.5	max	1216	–	–
13	Opus 5.5	medium	1213	–	–
14	Gemini 3.1 Pro	high	1209	–	–
15	GPT-5.2	xhigh	1207	–	–
16	Gemini 3 Pro	high	1206	–	–
17	Grok 4.20	xhigh	1205	–	–
18	Opus 4.8	max	1204	–	–
19	GPT-5.1	high	1200	–	–
20	GPT-5.4	xhigh	1199	–	–
21	Gemini 3 Flash	high	1196	–	–
22	Sonnet 5	max	1191	–	–
23	Opus 4.5	high	1180	–	–
24	GPT-6 Sol	max	1175	–	–
25	Grok 4.1 Fast	high	1171	–	–
26	Grok 4 Fast	none	1170	–	–
27	Grok 4.3	xhigh	1166	–	–
28	Sonnet 4.5	high	1160	–	–
29	GPT-6 Luna	max	1160	–	–
30	Opus 4.1	high	1148	–	–
31	o3	high	1145	–	–
32	Gemini 2.5 Pro	high	1143	–	–
33	Grok 4	high	1142	–	–
34	GPT-5	high	1133	–	–
35	Opus 4	medium	1127	–	–
36	GPT-4o	high	1007	–	–