Design Arena Leaderboard

312 models across 17 categories · 5.3M+ votes · Updated Jul 25, 2026

Crowdsourced design benchmark from designarena.ai — ELO ratings based on real user preferences

SVG

100 models · 160,087 votes

#	Model	Provider	ELO ▼	Win%	Battles	CI 95%	Avg Time
1	Claude Fable 5	anthropic	1352	69.0%	911	±12.7	54.1s
2	Arrow-1.0	quiver	1347	72.9%	3,093	±7.2	1m58s
3	Gemini 3.1 Pro Preview	google	1337	69.7%	4,022	±6.1	2m50s
4	Arrow 1.1 Max	quiver	1316	66.9%	2,273	±8.0	1m44s
5	Gemini 3.5 Flash	google	1296	62.2%	2,155	±7.9	1m17s
6	Gemini 3 Pro Preview	google	1295	72.1%	3,986	±6.4	48.7s
7	Claude Opus 4.6 (Thinking)	anthropic	1277	62.0%	3,704	±6.1	1m9s
8	Muse Spark	meta	1277	65.4%	2,664	±7.3	1m7s
9	Recraft V4.1 Vector Pro	recraft	1276	60.5%	2,996	±6.7	23.2s
10	Claude Opus 4.7	anthropic	1275	61.2%	2,817	±6.9	36.8s
11	GPT-5.5	openai	1274	59.2%	1,951	±8.2	55.8s
12	GLM 5.1	zhipu	1271	60.0%	2,646	±7.1	2m46s
13	Qwen3.7 Max	alibaba	1268	59.6%	2,774	±6.9	1m39s
14	Recraft V4 Vector	recraft	1266	60.1%	7,836	±4.3	25.8s
15	Claude Opus 4.6	anthropic	1265	60.7%	4,542	±5.5	55.6s
16	Recraft V4 Vector Pro	recraft	1264	59.0%	7,390	±4.4	33.5s
17	GLM 5.2	zhipu	1259	55.3%	642	±14.0	3m48s
18	Nex N2 Pro	nex-agi	1259	54.1%	1,291	±9.8	3m26s
19	GLM 5 Turbo	zhipu	1258	57.8%	3,292	±6.3	2m36s
20	Claude Sonnet 4.6	anthropic	1247	59.1%	3,923	±5.8	55.9s
21	GPT-5.4 (Medium)	openai	1244	57.8%	3,399	±6.2	1m5s
22	Claude Sonnet 5	anthropic	1241	53.8%	1,639	±8.8	50.2s
23	GPT-5 (High)	openai	1238	64.1%	1,688	±9.1	2m14s
24	Arrow 1.1	quiver	1236	57.0%	2,518	±7.3	1m4s
25	Claude Opus 4.5	anthropic	1230	58.7%	4,790	±5.3	31.9s
26	Kimi K2.6	moonshot	1230	52.4%	2,897	±6.7	2m4s
27	Claude Opus 4.8	anthropic	1226	53.4%	1,852	±8.3	26.2s
28	MiniMax M3	minimax	1226	51.3%	1,539	±9.0	1m10s
29	Kimi K2.7 Code	moonshot	1224	49.2%	827	±12.2	2m21s
30	MiMo-V2.5-Pro	xiaomi	1224	52.3%	885	±12.0	2m7s
31	MiMo-V2.5	xiaomi	1220	52.6%	1,708	±8.6	1m14s
32	GLM 5	zhipu	1217	54.4%	7,455	±4.3	57.1s
33	Grok 4.20 Beta (Reasoning)	xai	1214	53.8%	3,798	±5.9	1m14s
34	Qwen3.6 Plus	alibaba	1210	51.6%	4,821	±5.2	2m22s
35	DeepSeek-V4-Flash	deepseek	1204	48.6%	1,968	±8.1	1m51s
36	Claude Opus 4.1 (Thinking)	anthropic	1202	60.8%	1,526	±9.5	33.0s
37	GLM 5V Turbo	zhipu	1201	51.1%	4,141	±5.6	42.9s
38	MiniMax M2.5	minimax	1200	54.5%	1,238	±10.2	46.7s
39	GPT-5.1 (High)	openai	1199	57.4%	2,721	±7.0	1m15s
40	Kimi K2.5 (Thinking)	moonshot	1194	48.1%	1,486	±9.2	1m47s
41	GLM 4.7	zhipu	1192	54.3%	7,652	±4.2	1m30s
42	GPT-5.4 (Low)	openai	1191	53.2%	3,848	±5.9	23.4s
43	Qwen3.5 397B A17B	alibaba	1191	56.1%	969	±11.7	1m13s
44	MiniMax M2.7	minimax	1188	50.4%	3,821	±5.9	53.6s
45	GPT-5.2 (Medium)	openai	1188	53.7%	5,058	±5.2	44.5s
46	DeepSeek-V4-Pro	deepseek	1186	46.6%	1,724	±8.6	4m17s
47	GPT-5.2 (Low)	openai	1185	53.1%	4,777	±5.3	40.5s
48	GPT-5.1 (Medium)	openai	1181	55.4%	3,304	±6.4	37.8s
49	MiniMax M2.1	minimax	1181	55.4%	2,075	±8.0	40.7s
50	GPT-5.2 (None)	openai	1180	54.2%	6,010	±4.7	40.2s
51	GPT-5.3 Codex	openai	1179	54.0%	3,022	±6.6	17.6s
52	Claude Opus 4	anthropic	1177	57.7%	1,732	±8.9	22.2s
53	GPT-5 (Minimal)	openai	1177	52.1%	4,702	±5.3	59.8s
54	GPT-5.4 (None)	openai	1175	50.4%	3,894	±5.8	20.4s
55	GPT-5.1 (Low)	openai	1167	55.6%	3,030	±6.7	32.1s
56	Claude Sonnet 4.5 (Thinking)	anthropic	1164	52.4%	5,043	±5.2	26.4s
57	Claude Opus 4.1	anthropic	1159	50.1%	4,944	±5.2	27.5s
58	GLM 4.6	zhipu	1158	52.1%	1,920	±8.3	33.2s
59	Claude Sonnet 4.5	anthropic	1157	49.3%	4,804	±5.2	27.4s
60	Qwen3.5 Plus 02-15	alibaba	1157	48.9%	3,354	±6.3	1m5s
61	GPT-5.1 (None)	openai	1154	50.6%	3,075	±6.6	31.6s
62	MiniMax M2 Stable	minimax	1149	55.3%	1,575	±9.1	32.6s
63	GLM 4.5	zhipu	1146	50.8%	1,961	±8.2	36.6s
64	GPT-5 mini (Default)	openai	1140	45.9%	4,696	±5.3	48.7s
65	Claude 3.7 Sonnet	anthropic	1137	52.1%	1,850	±8.4	18.8s
66	Grok 4.3	xai	1132	41.2%	2,307	±7.6	32.2s
67	Nemotron 3 Ultra	nvidia	1129	37.8%	1,160	±10.7	46.2s
68	Claude Sonnet 4 (Thinking)	anthropic	1128	51.1%	1,788	±8.6	21.0s
69	Step 3.7 Flash	stepfun	1122	39.5%	1,413	±9.7	57.8s
70	GLM 4.5 Air	zhipu	1121	50.8%	1,639	±9.0	41.9s
71	DeepSeek-V3.1 (Thinking)	deepseek	1114	50.1%	1,684	±8.8	1m20s
72	Claude Sonnet 4	anthropic	1112	47.8%	1,724	±8.7	17.0s
73	v0-1.5-md	vercel	1111	48.1%	1,217	±10.3	22.6s
74	Grok 4.20 Beta	xai	1106	41.7%	3,919	±5.9	16.5s
75	Gemini 3.1 Flash-Lite Preview	google	1101	42.5%	3,757	±6.0	3.4s
76	DeepSeek-R1-0528	deepseek	1088	48.7%	1,568	±9.2	1m0s
77	GLM 4.7 Flash	zhipu	1087	44.2%	1,587	±9.0	1m22s
78	DeepSeek-V3.2	deepseek	1082	40.9%	4,615	±5.5	47.9s
79	DeepSeek-V3.2-Exp	deepseek	1081	42.0%	2,352	±7.6	54.9s
80	MiMo-V2-Flash	xiaomi	1079	40.5%	4,800	±5.4	18.8s
81	Claude Haiku 4.5	anthropic	1075	39.1%	4,717	±5.5	11.6s
82	Grok 4	xai	1071	41.4%	2,454	±7.5	36.8s
83	Gemini 2.5 Flash Preview 09-2025	google	1069	43.1%	1,789	±8.7	17.5s
84	Trinity Large Thinking	arceeai	1065	35.2%	2,774	±7.2	42.7s
85	Qwen3 Max	alibaba	1060	37.2%	4,185	±5.8	31.4s
86	Grok 4.1 Fast (Reasoning)	xai	1058	37.6%	4,530	±5.6	34.6s
87	Mistral Large 3 (2512)	mistral	1042	38.0%	4,911	±5.4	25.0s
88	Mistral Medium 3.1 (2508)	mistral	1042	38.2%	4,843	±5.4	23.0s
89	Prime Intellect: INTELLECT-3	primeintellect	1041	34.4%	4,479	±5.7	27.1s
90	Grok 3	xai	1033	36.1%	2,964	±7.0	11.9s
91	GPT-5.1 Codex Mini	openai	1027	35.3%	4,888	±5.5	13.8s
92	Mercury 2	inception	1027	25.2%	854	±13.8	2.4s
93	DeepSeek-V3-0324	deepseek	1026	38.8%	1,708	±9.1	15.2s
94	Grok 4 Fast (Reasoning)	xai	1025	34.9%	5,025	±5.4	12.4s
95	DeepSeek-V3.1	deepseek	1016	38.2%	1,825	±8.8	7.8s
96	Mistral Medium 3.5	mistral	997	26.8%	1,159	±11.8	11.8s
97	Grok 4.1 Fast	xai	992	31.5%	4,511	±5.8	10.7s
98	Grok 4 Fast	xai	988	30.0%	4,863	±5.7	7.6s
99	AesCoder-4B	designflow	980	27.3%	5,185	±5.6	20.7s
100	Olmo 3.1 32B Think	allenai	921	25.9%	1,852	±9.5	1m38s