Leaderboard

#	Model	ELO	Peak	Avg Score	Avg Cost	Consistency
1	Claude Opus 4.7	1904	1904	88.3	$1.47	100.0%
2	GPT 5.5	1867	1868	86.8	$1.61	94.3%
3	GPT 5.4 Mini	1791	1792	84.3	$0.38	95.7%
4	Claude Opus 4.6	1782	1782	84.7	$1.11	97.1%
5	Claude Sonnet 4.6	1770	1771	84.3	$0.24	93.8%
6	Claude Opus 4.5	1722	1723	82.3	$0.88	92.9%
7	Qwen3.7 Max	1715	1717	80.2	$1.00	92.7%
8	Kimi K2.6	1704	1705	80.8	$0.47	91.4%
9	GLM 5.1	1695	1696	80.7	$0.22	92.9%
10	Deepseek V4 Pro	1684	1685	80.2	$0.37	90.0%
11	GPT 5.2	1673	1674	80.2	$0.19	84.4%
12	GPT 5.2 Codex	1671	1672	78.5	$0.13	80.3%
13	GPT 5.3 Codex Spark	1670	1672	79.3	$0.32	92.9%
14	Deepseek V4 Flash	1668	1669	79.6	$0.06	88.6%
15	Gemini 3.5 Flash	1666	1667	79.8	$0.27	94.3%
16	GPT 5.3 Codex	1666	1667	80.2	$0.13	82.8%
17	Minimax M2.7 [NVFP4]	1648	1649	79.4	<$0.01	92.9%
18	Qwen3.6 Plus	1639	1640	77.8	$0.40	90.0%
19	GPT 5.1 Codex Mini	1624	1625	76.6	$0.59	82.2%
20	Qwen3.6 35b A3b Q4 K XL [Q4_K_XL]	1609	1610	76.5	<$0.01	85.9%
21	Gemini 3.1 Pro Preview	1590	1591	75.9	$0.57	70.3%
22	Claude Sonnet 4.5	1583	1583	76.1	$0.25	72.3%
23	Minimax M2.7	1582	1583	76.4	$0.05	78.6%
24	Qwen3.6 27b [Q4_K_XL]	1581	1618	75.1	<$0.01	78.6%
25	Gemini 3 Pro Preview	1569	1570	74.9	$0.46	72.3%
26	Qwen3.6 35b A3b [BF16]	1559	1560	74.7	<$0.01	72.9%
27	Qwen3.5 397b A17b Q4 K XL [Q4_K_XL]	1553	1554	74.8	$0.85	75.7%
28	GLM 5	1551	1553	73.2	$0.15	69.4%
29	Claude Haiku 4.5	1523	1524	71.5	$0.07	63.1%
30	Qwen3.5 122b A10b [Q4_K_XL]	1520	1521	71.8	$0.24	60.2%
31	Kimi K2.5	1520	1521	72.4	$0.06	68.9%
32	Gemini 3 Flash Preview	1515	1516	72.4	$0.02	67.4%
33	GLM 4.7	1514	1515	71.8	$0.10	64.5%
34	Qwen3.5 Plus 02.15	1510	1511	70.6	$0.13	56.5%
35	Grok 4	1507	1508	71.8	$0.27	66.9%
36	Qwen3.6 27b [BF16]	1493	1494	70.4	<$0.01	57.1%
37	GLM 4.7 [Q4_K_XL]	1478	1479	71.2	$0.04	56.1%
38	Qwen3.5 27b	1446	1447	70.0	$0.36	59.8%
39	Qwen3.5 122b A10b	1443	1444	69.6	$0.38	57.3%
40	Gemini 2.5 Pro	1437	1439	68.4	$0.27	53.4%
41	Grok 4.1 Fast	1424	1425	68.7	$0.05	60.2%
42	Minimax M2.1	1419	1420	65.1	$0.05	46.4%
43	Deepseek R1 0528	1416	1417	64.6	$0.05	43.7%
44	Deepseek V3.2	1389	1390	64.0	$0.04	37.5%
45	Qwen3 Coder	1383	1385	60.8	$0.11	37.8%
46	GLM 4.6	1383	1384	64.4	$0.11	40.7%
47	Minimax M2.5	1383	1384	65.5	$0.14	45.5%
48	Qwen3 Coder Plus	1378	1379	63.2	$0.07	39.3%
49	GLM 4.5	1378	1380	65.9	$0.10	49.5%
50	Grok Code Fast 1	1376	1377	65.8	$0.07	42.4%
51	Qwen3.5 35b A3b	1376	1377	63.2	$0.08	42.1%
52	Minimax M2.5 [Q4_K_XL]	1373	1374	63.9	$0.03	33.3%
53	Qwen3.5 Flash 02.23	1368	1369	63.4	$0.06	46.3%
54	Qwen3.5 27b [Q4_K_M]	1363	1364	63.1	$0.18	42.7%
55	Devstral 2512	1352	1353	63.0	$0.10	32.4%
56	Step 3.5 Flash	1338	1339	60.4	$0.13	39.8%
57	Qwen3 Coder Next	1328	1329	61.2	$0.02	36.1%
58	GLM 4.5 Air	1327	1328	60.3	$0.03	29.1%
59	GPT OSS 120b	1322	1323	59.1	$0.12	30.2%
60	Trinity Large Preview:free	1303	1304	54.5	$0.05	20.3%
61	Qwen3 Coder Flash	1294	1295	59.1	$0.02	27.5%
62	Qwen3 Coder Next [Q4_K_XL]	1292	1293	58.3	<$0.01	19.5%
63	Gemini 2.5 Flash Lite	1281	1282	56.6	$0.02	21.5%
64	Qwen3.5 35b A3b [Q4_K_XL]	1249	1250	49.4	$0.05	17.2%
65	GPT OSS 20b	1246	1247	53.8	$0.11	21.9%
66	GLM 4.7 Flash	1236	1237	55.8	$0.01	22.4%
67	Nemotron 3 Nano 30b A3b	1217	1218	52.3	$0.09	18.3%
68	Qwen3 Coder 30b [Q4_K_M]	1163	1164	52.1	$0.01	12.5%

Leaderboard

ELO Distribution