3D generation rankings

Best AI for 3D

Compare models on Design Arena 3D tournaments. Preference ELO stays task-specific and does not rewrite the Vibe Coding Index.

Current source snapshot

Jul 19, 2026

121 ranked entries

Live category evidence

Current Design Arena leaderboard

How this source works →

Rank	Model	ELO	Win rate	Average time	Blended price	Tournament runs
1	Kimi K3 Moonshot	1,461	73.3%	621.3s	$6	674
2	Claude Fable 5 Anthropic	1,365	66.6%	129.2s	$22	9,212
3	GLM 5.2 Zhipu	1,363	62.3%	412.0s	$2.15	11,609
4	Grok 4.5 xAI	1,346	61.3%	92.0s	$3	8,775
5	Claude Opus 4.6 (Thinking) Anthropic	1,342	64.0%	153.6s	$11	11,891
6	Kimi K2.6 Moonshot	1,341	62.1%	331.3s	$1.71	28,760
7	Claude Opus 4.6 Anthropic	1,340	63.6%	156.0s	$11	19,768
9	Qwen3.7 Plus Alibaba	1,322	55.6%	190.3s	$0.56	956
10	Claude Sonnet 5 Anthropic	1,321	57.0%	165.2s	$4	5,230
11	GLM 5.1 Zhipu	1,320	59.8%	322.0s	$2.14	6,832
12	DeepSeek-V4-Pro DeepSeek	1,319	59.3%	329.5s	$2.18	10,272
13	Qwen3.7 Max Alibaba	1,319	58.7%	187.6s	$2.21	13,642
14	GLM 5 Turbo Zhipu	1,317	59.3%	308.0s	$1.9	12,894
15	MiMo-V2.5-Pro Xiaomi	1,312	57.3%	203.0s	$1.5	5,091
16	Claude Opus 4.7 Anthropic	1,311	58.8%	97.3s	$11	19,576
17	Muse Spark 1.1 Meta	1,310	54.6%	199.0s	$2	13,299
18	Kimi K2.7 Code Moonshot	1,310	56.0%	257.3s	$1.71	11,579
19	Claude Sonnet 4.6 Anthropic	1,300	58.4%	132.3s	$6	20,333
20	Gemini 3.1 Pro Preview Google	1,298	60.3%	89.6s	$4.5	36,401
21	Gemini 3.5 Flash Google	1,297	57.7%	60.1s	$3.38	29,829
22	Nex N2 Pro Nex Agi	1,296	53.1%	477.3s	$0.44	3,755
23	GLM 5 Zhipu	1,295	56.4%	134.4s	$1.55	32,017
24	MiniMax M3 MiniMax	1,287	55.4%	123.2s	$0.53	7,461
25	Gemini 3 Pro Preview Google	1,287	66.4%	72.7s	$4.5	12,275
26	MiMo-V2.5 Xiaomi	1,281	52.9%	122.9s	$0.14	14,558
27	Claude Opus 4.5 Anthropic	1,280	58.6%	86.1s	$10	18,998
28	Claude Opus 4.8 Anthropic	1,279	55.7%	70.8s	$10	16,311
29	GLM 5V Turbo Zhipu	1,276	54.8%	106.0s	$1.9	29,832
30	Kimi K2.5 (Thinking) Moonshot	1,271	53.4%	148.5s	$1.3	27,147
31	Hy3 Tencent	1,267	49.9%	259.8s	$0.25	1,047
32	Qwen3.6 Plus Alibaba	1,262	51.4%	167.8s	$0.73	16,566
33	GPT-5.5 OpenAI	1,258	53.0%	145.2s	$11.25	23,121
34	GLM 4.7 Zhipu	1,257	54.3%	144.2s	$1	29,268
35	Grok 4.20 Beta (Reasoning) xAI	1,257	53.7%	96.4s	$1.56	17,038
36	MiniMax M2.7 MiniMax	1,256	50.8%	132.7s	$0.53	14,675
37	DeepSeek-V4-Flash DeepSeek	1,256	49.8%	103.1s	$0.17	11,380
38	Gemini 3 Flash Preview Google	1,250	62.7%	24.4s	$1.13	9,357
39	GLM 4.5 Zhipu	1,237	59.7%	108.7s	$1	9,161
40	MiniMax M2.5 MiniMax	1,236	57.6%	93.2s	$0.44	5,389
41	MiniMax M2.1 MiniMax	1,231	57.5%	96.1s	$0.53	9,381
42	Qwen3.5 397B A17B Alibaba	1,227	56.7%	104.7s	$1.09	4,149
43	Claude Sonnet 4.5 Anthropic	1,225	52.5%	66.2s	$6.6	21,364
44	Claude Sonnet 4.5 (Thinking) Anthropic	1,218	51.6%	62.0s	$6.6	20,943
45	Claude Opus 4.1 Anthropic	1,217	51.6%	76.8s	$30	20,472
46	DeepSeek-V3.2-Exp DeepSeek	1,215	56.4%	152.4s	$0.31	9,583
47	Grok 4.20 Beta xAI	1,215	48.3%	49.7s	$1.56	17,353
48	Claude Opus 4.1 (Thinking) Anthropic	1,207	56.5%	76.9s	$30	7,473
49	DeepSeek-V3.1 (Thinking) DeepSeek	1,207	56.0%	155.7s	$0.44	7,784
50	Claude Sonnet 4 Anthropic	1,206	57.8%	57.0s	$6	8,316
51	Claude Opus 4 Anthropic	1,205	57.7%	66.6s	$30	7,929
52	DeepSeek-V3.2 DeepSeek	1,200	49.6%	132.9s	$1	14,620
53	Nemotron 3 Ultra NVIDIA	1,200	42.9%	119.2s	$1.05	4,921
54	GLM 4.6 Zhipu	1,194	54.0%	80.5s	$0.96	14,578
55	MiMo-V2-Flash Xiaomi	1,193	49.2%	44.7s	—	16,624
56	Claude Sonnet 4 (Thinking) Anthropic	1,192	54.9%	53.6s	$6	7,680
57	Grok 4.3 xAI	1,191	44.2%	52.8s	$1.56	29,451
58	GLM 4.5 Air Zhipu	1,189	54.1%	86.6s	$0.43	8,103
59	Hy3 Low Tencent	1,189	38.9%	57.8s	$0.25	1,268
60	GLM 4.7 Flash Zhipu	1,187	51.2%	116.9s	$0.15	5,798
61	Step 3.7 Flash Stepfun	1,186	42.5%	96.5s	$0.44	8,234
62	Qwen3.5 Plus 02-15 Alibaba	1,185	47.8%	97.7s	$0.59	9,623
63	DeepSeek-R1-0528 DeepSeek	1,178	53.4%	68.6s	$1.15	8,357
64	Claude 3.7 Sonnet Anthropic	1,172	51.9%	79.3s	—	7,529
65	GPT-5.4 (Medium) OpenAI	1,167	42.3%	196.1s	$5.63	10,752
66	Mistral Large 3 (2512) Mistral	1,165	46.9%	66.8s	$0.75	14,978
67	Prime Intellect. INTELLECT-3 Prime Intellect	1,164	45.1%	59.1s	—	15,077
68	MiniMax M2 Stable MiniMax	1,164	48.3%	101.4s	$0.45	5,275
69	DeepSeek-V3-0324 DeepSeek	1,154	50.7%	43.2s	$0.63	9,090
70	Gemini 2.5 Pro Google	1,149	52.2%	56.9s	$3.44	3,849
71	Mistral Medium 3 (2505) Mistral	1,149	54.6%	51.4s	$0.8	2,890
72	Mistral Medium 3.1 (2508) Mistral	1,148	44.7%	77.9s	$0.8	14,173
73	Trinity Large Thinking Arcee Ai	1,147	41.3%	67.7s	$0.39	6,482
74	Grok Code Fast 1 xAI	1,146	50.4%	13.7s	—	8,460
75	GPT-5.2 (Medium) OpenAI	1,146	41.6%	138.8s	$4.81	15,472
76	DeepSeek-V3.1 DeepSeek	1,144	48.0%	28.8s	$0.46	9,535
77	Claude Haiku 4.5 Anthropic	1,139	41.2%	29.8s	$2.2	18,482
78	Grok 4.1 Fast xAI	1,139	42.8%	24.3s	—	16,318
79	GPT-5.2 (None) OpenAI	1,139	40.8%	131.7s	$4.81	16,148
80	Qwen3 Max Alibaba	1,139	43.5%	73.0s	$1.56	16,354
81	Gemini 2.5 Flash Preview 09-2025 Google	1,137	47.4%	22.4s	$0.85	9,097
82	GPT-5.4 (Low) OpenAI	1,133	39.2%	65.1s	$5.63	11,777
83	Grok 4 Fast xAI	1,133	41.7%	31.5s	—	18,019
84	GPT-5.2 (Low) OpenAI	1,133	42.1%	127.4s	$4.81	13,533
85	Grok 4 Fast (Reasoning) xAI	1,129	41.2%	21.0s	—	18,584
86	GPT-5.1 (High) OpenAI	1,128	43.9%	150.1s	$3.44	8,405
87	Grok 4 xAI	1,128	44.3%	52.6s	—	11,878
88	Grok 3 xAI	1,127	44.3%	27.2s	—	13,192
89	GPT-5 (Minimal) OpenAI	1,122	41.4%	117.3s	$3.44	17,333
90	GPT-5.2 (High) OpenAI	1,120	45.3%	120.0s	$4.81	1,661
91	GPT-5.1 (None) OpenAI	1,117	43.0%	81.7s	$3.44	10,899
92	GPT-5.4 (None) OpenAI	1,115	37.1%	55.0s	$5.63	13,539
93	Gemini 3.1 Flash-Lite Preview Google	1,114	38.8%	4.5s	$0.56	12,911
94	Devstral Medium Mistral	1,110	48.1%	17.0s	—	3,285
95	GPT-5.1 (Low) OpenAI	1,110	42.1%	74.6s	$3.44	10,832
96	GPT-5.1 (Medium) OpenAI	1,105	41.5%	95.2s	$3.44	10,709
97	GPT-5 mini (Default) OpenAI	1,104	37.0%	101.0s	$0.69	19,372
98	Grok 4.1 Fast (Reasoning) xAI	1,100	37.4%	57.7s	—	16,053
99	Ministral 3 8B (2512) Mistral	1,098	46.2%	33.2s	$0.3	1,034
100	GPT-5 (High) OpenAI	1,095	40.2%	155.4s	$3.44	6,807
101	Gemini 2.5 Flash Google	1,088	44.6%	45.2s	$0.85	3,254
102	Codestral 2508 Mistral	1,087	45.5%	7.8s	$0.45	3,142
103	Olmo 3.1 32B Think Allenai	1,079	35.5%	109.3s	—	8,236
104	GPT-5.3 Codex OpenAI	1,074	35.3%	46.0s	$4.81	11,788
105	Qwen3-235B-A22B-Thinking-2507 Alibaba	1,065	40.7%	87.6s	$0.75	2,908
106	Llama 3.1 Nemotron Ultra 253B NVIDIA	1,061	39.0%	21.5s	—	1,464
107	Qwen3-235B-A22B-Instruct-2507 Alibaba	1,060	41.1%	39.5s	$0.3	3,206
108	Ministral 3 14B (2512) Mistral	1,056	39.6%	53.6s	$0.35	1,017
109	GPT-5.1 Codex Mini OpenAI	1,055	32.9%	34.4s	$0.69	17,161
110	Mercury 2 Inception	1,050	24.9%	3.3s	$0.38	3,943
111	Gemini 2.5 Flash Lite Preview 09-2025 Google	1,034	36.5%	6.5s	—	3,139
112	Grok 3 Mini xAI	1,033	41.5%	18.4s	—	3,594
113	GPT-5 nano (Default) OpenAI	1,030	36.1%	88.2s	$0.14	3,043
114	Ministral 3 3B (2512) Mistral	1,029	35.9%	29.2s	$0.15	1,253
115	GPT-4.1 nano OpenAI	991	46.0%	17.1s	$0.18	858
116	GPT OSS 120B OpenAI	967	29.4%	16.9s	$0.26	2,487
117	Llama 4 Maverick Meta Llama	966	40.2%	9.5s	$0.35	790
118	GPT-4o OpenAI	935	39.2%	11.5s	$4.38	850
119	Qwen3-235B-A22B Alibaba	925	24.5%	67.8s	$0.8	2,399
120	o4-mini OpenAI	919	34.0%	30.6s	$1.93	885
121	GPT-4.1 OpenAI	916	30.9%	67.8s	$3.5	763
122	GPT-4.1 mini OpenAI	903	30.5%	61.7s	$0.7	688

What this ranking means. ELO summarizes head-to-head preference results inside this category. It is useful evidence for the named task, but it remains separate from the broader Vibe Coding Index.