Design Arena Leaderboard

432 models across 23 categories · 8.0M+ votes · Updated Sep 23, 2026

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

SVG

119 models · 184,708 votes

#	Model	Provider	ELO ▼	Win%	Battles	CI 95%	Avg Time
1	GPT-6 Astra (xhigh)	openai	1485	78.4%	695	±16.3	4m13s
2	Muse Spark 1.3 (xhigh)	meta	1343	62.1%	660	±14.4	1m18s
3	Claude Fable 5.1	anthropic	1342	63.3%	908	±12.3	1m16s
4	Claude Opus 5	anthropic	1340	61.5%	1,143	±10.8	1m46s
5	GPT-5.6 Sol (xhigh)	openai	1333	59.9%	678	±14.0	1m56s
6	Kimi K3	moonshot	1325	62.3%	745	±13.4	6m30s
7	GPT-5.6 Sol (Medium)	openai	1324	61.8%	1,354	±10.0	46.8s
8	Claude Fable 5	anthropic	1323	63.7%	1,831	±8.7	58.1s
9	GLM-5.3	zhipu	1315	58.8%	1,633	±9.0	4m22s
10	Muse Spark 1.2	meta	1308	56.4%	1,641	±9.0	2m46s
11	GLM-5.3-Flash	zhipu	1308	57.1%	1,948	±8.3	3m22s
12	Gemini 3.1 Pro Preview	google	1305	67.8%	4,620	±5.7	2m48s
13	Arrow-1.0	quiver	1293	66.4%	4,081	±6.0	1m57s
14	Arrow 1.1 Max	quiver	1290	62.7%	3,479	±6.3	1m57s
15	Arrow 2	quiver	1290	55.8%	1,365	±9.9	1m38s
16	Muse Spark 1.1	meta	1274	52.2%	1,920	±8.2	2m12s
17	Gemini 3.5 Flash	google	1274	60.0%	2,932	±6.8	1m23s
18	Gemini 3 Pro Preview	google	1264	72.1%	3,987	±6.4	48.7s
19	Grok 4.6	xai	1258	50.5%	1,090	±10.8	4m2s
20	GPT-5.5	openai	1255	57.2%	2,802	±6.8	54.0s
21	Claude Opus 4.6 (Thinking)	anthropic	1247	60.9%	3,987	±5.8	1m16s
22	Muse Spark	meta	1246	65.4%	2,664	±7.3	1m7s
23	Recraft V4.1 Vector Pro	recraft	1246	59.6%	3,371	±6.3	22.5s
24	GLM 5.2	zhipu	1242	50.1%	1,959	±8.1	2m57s
25	GPT-5.6 Terra	openai	1241	51.7%	5,924	±4.9	22.2s
26	Qwen3.7 Max	alibaba	1239	57.6%	3,328	±6.3	1m39s
27	Recraft V4 Vector	recraft	1239	59.9%	8,406	±4.1	25.5s
28	Grok 4.5	xai	1238	48.6%	2,331	±7.5	1m3s
29	GLM 5.1	zhipu	1237	56.5%	3,587	±6.1	3m1s
30	Recraft V4 Vector Pro	recraft	1237	58.7%	7,971	±4.2	33.0s
31	Claude Opus 4.6	anthropic	1234	60.5%	4,571	±5.5	55.7s
32	Qwen3.7 Plus	alibaba	1230	49.1%	1,051	±11.0	1m58s
33	GLM 5 Turbo	zhipu	1226	55.5%	4,165	±5.6	2m50s
34	Hy3	tencent	1222	46.0%	827	±12.5	3m11s
35	Claude Sonnet 4.6	anthropic	1216	58.7%	4,023	±5.8	56.4s
36	Nex N2 Pro	nex-agi	1216	50.6%	2,001	±7.9	2m57s
37	GPT-5.6 Luna	openai	1215	48.2%	1,025	±11.2	18.2s
38	GPT-5.4 (Medium)	openai	1214	57.8%	3,398	±6.2	1m5s
39	Arrow 1.1	quiver	1213	55.1%	3,224	±6.4	1m5s
40	Claude Opus 4.8	anthropic	1207	52.1%	2,447	±7.2	30.2s
41	Claude Sonnet 5	anthropic	1207	50.9%	2,129	±7.7	53.8s
42	DeepSeek-V4-Flash-0731	deepseek	1207	45.2%	702	±13.5	3m39s
43	GPT-5 (High)	openai	1207	64.1%	1,688	±9.1	2m14s
44	MiMo-V2.5-Pro	xiaomi	1205	49.8%	1,515	±9.2	2m2s
45	Kimi K2.7 Code	moonshot	1203	47.2%	1,461	±9.3	2m29s
46	Kimi K2.6	moonshot	1201	51.2%	3,539	±6.0	2m7s
47	Claude Opus 4.5	anthropic	1199	58.7%	4,785	±5.3	31.9s
48	MiMo-V2.5	xiaomi	1196	50.9%	2,199	±7.6	1m27s
49	GLM 5	zhipu	1187	54.4%	7,455	±4.3	57.0s
50	MiniMax M3	minimax	1186	47.5%	2,246	±7.5	59.3s
51	Grok 4.20 Beta (Reasoning)	xai	1184	52.4%	4,286	±5.5	1m19s
52	Qwen3.6 Plus	alibaba	1181	50.9%	5,193	±5.0	2m22s
53	DeepSeek-V4-Flash	deepseek	1175	48.4%	1,990	±8.0	1m52s
54	Claude Opus 4.1 (Thinking)	anthropic	1171	60.8%	1,526	±9.5	33.0s
55	GLM 5V Turbo	zhipu	1170	49.6%	4,754	±5.2	46.1s
56	Kimi K2.5 (Thinking)	moonshot	1169	48.4%	1,549	±9.1	1m47s
57	MiniMax M2.5	minimax	1168	54.5%	1,238	±10.3	46.7s
58	GPT-5.1 (High)	openai	1168	57.4%	2,722	±7.1	1m15s
59	GLM 4.7	zhipu	1162	54.3%	7,652	±4.2	1m30s
60	GPT-5.4 (Low)	openai	1161	53.2%	3,842	±5.9	23.4s
61	DeepSeek-V4-Pro	deepseek	1159	45.4%	2,058	±7.9	4m5s
62	Qwen3.5 397B A17B	alibaba	1158	56.1%	969	±11.7	1m13s
63	GPT-5.2 (Medium)	openai	1158	53.7%	5,056	±5.2	44.5s
64	MiniMax M2.7	minimax	1156	48.3%	4,402	±5.4	53.7s
65	GPT-5.2 (Low)	openai	1154	53.1%	4,772	±5.3	40.5s
66	GPT-5.1 (Medium)	openai	1150	55.4%	3,304	±6.4	37.8s
67	MiniMax M2.1	minimax	1150	55.4%	2,075	±8.0	40.7s
68	GPT-5.2 (None)	openai	1149	54.1%	6,005	±4.8	40.1s
69	GPT-5.3 Codex	openai	1148	54.0%	3,023	±6.6	17.6s
70	Claude Opus 4	anthropic	1146	57.7%	1,732	±8.9	22.2s
71	GPT-5 (Minimal)	openai	1146	52.0%	4,698	±5.4	59.7s
72	GPT-5.4 (None)	openai	1145	50.4%	3,892	±5.8	20.4s
73	GPT-5.1 (Low)	openai	1137	55.6%	3,030	±6.7	32.1s
74	Claude Sonnet 4.5 (Thinking)	anthropic	1134	52.2%	5,117	±5.2	26.7s
75	Inkling	thinkingmachines	1134	35.5%	843	±12.8	25.6s
76	Claude Opus 4.1	anthropic	1128	50.1%	4,945	±5.2	27.5s
77	Claude Sonnet 4.5	anthropic	1127	49.0%	4,913	±5.2	27.6s
78	GLM 4.6	zhipu	1127	52.1%	1,920	±8.3	33.2s
79	Qwen3.5 Plus 02-15	alibaba	1126	48.9%	3,352	±6.3	1m5s
80	GPT-5.1 (None)	openai	1123	50.6%	3,075	±6.6	31.6s
81	MiniMax M2 Stable	minimax	1119	55.3%	1,575	±9.2	32.6s
82	GLM 4.5	zhipu	1115	50.8%	1,961	±8.2	36.6s
83	GPT-5 mini (Default)	openai	1110	45.8%	4,694	±5.4	48.7s
84	Claude 3.7 Sonnet	anthropic	1107	52.1%	1,850	±8.4	18.8s
85	Grok 4.3	xai	1104	39.0%	2,968	±6.8	28.6s
86	Claude Sonnet 4 (Thinking)	anthropic	1097	51.1%	1,788	±8.6	21.0s
87	Nemotron 3 Ultra	nvidia	1095	35.3%	1,720	±8.9	41.7s
88	Step 3.7 Flash	stepfun	1095	37.3%	1,963	±8.3	1m3s
89	GLM 4.5 Air	zhipu	1090	50.8%	1,639	±9.0	41.9s
90	DeepSeek-V3.1 (Thinking)	deepseek	1084	50.1%	1,684	±8.8	1m20s
91	Claude Sonnet 4	anthropic	1082	47.8%	1,724	±8.8	17.0s
92	v0-1.5-md	vercel	1080	48.1%	1,217	±10.3	22.6s
93	Grok 4.20 Beta	xai	1076	40.4%	4,348	±5.6	17.8s
94	Gemini 3.1 Flash-Lite Preview	google	1071	42.5%	3,753	±6.0	3.4s
95	DeepSeek-R1-0528	deepseek	1057	48.7%	1,568	±9.2	1m0s
96	GLM 4.7 Flash	zhipu	1057	44.2%	1,587	±9.1	1m22s
97	DeepSeek-V3.2	deepseek	1051	40.8%	4,608	±5.5	47.8s
98	DeepSeek-V3.2-Exp	deepseek	1051	42.0%	2,352	±7.6	54.9s
99	MiMo-V2-Flash	xiaomi	1049	40.5%	4,801	±5.4	18.8s
100	Claude Haiku 4.5	anthropic	1044	39.1%	4,714	±5.5	11.6s
101	Grok 4	xai	1040	41.4%	2,454	±7.5	36.8s
102	Gemini 2.5 Flash Preview 09-2025	google	1039	43.1%	1,789	±8.7	17.5s
103	Trinity Large Thinking	arceeai	1035	35.2%	2,774	±7.2	42.7s
104	Qwen3 Max	alibaba	1030	37.2%	4,182	±5.8	31.4s
105	Grok 4.1 Fast (Reasoning)	xai	1028	37.5%	4,527	±5.6	34.6s
106	Mercury 2	inception	1013	24.9%	1,491	±10.5	2.9s
107	Mistral Large 3 (2512)	mistral	1012	38.0%	4,910	±5.5	25.0s
108	Mistral Medium 3.1 (2508)	mistral	1011	38.2%	4,843	±5.5	23.0s
109	Prime Intellect: INTELLECT-3	primeintellect	1010	34.4%	4,479	±5.7	27.1s
110	Grok 3	xai	1002	36.1%	2,964	±7.0	11.9s
111	GPT-5.1 Codex Mini	openai	997	35.3%	4,888	±5.5	13.8s
112	DeepSeek-V3-0324	deepseek	995	38.8%	1,708	±9.2	15.2s
113	Grok 4 Fast (Reasoning)	xai	995	34.9%	5,020	±5.5	12.4s
114	DeepSeek-V3.1	deepseek	986	38.2%	1,825	±8.8	7.8s
115	Mistral Medium 3.5	mistral	969	26.7%	1,157	±11.9	11.8s
116	Grok 4.1 Fast	xai	962	31.5%	4,504	±5.9	10.7s
117	Grok 4 Fast	xai	958	30.0%	4,865	±5.7	7.6s
118	AesCoder-4B	designflow	950	27.3%	5,179	±5.7	20.7s
119	Olmo 3.1 32B Think	allenai	891	25.9%	1,852	±9.5	1m38s